# 2.

1
TAB
6
7
9
10
11

TBC
2
3
2
3
2
3
2
3
2
3

CAB
1200
1200
550
550
550

CBC
300
300
300
300
300
300
300
300
300
300

TAC
8
9
9
10
11
12
12
13
13
14

CAC
1500
1500
1500
1500
850
850
850
850
850
850

(a) Sample space of travel time from A to B = {6, 7, 9, 10, 11}
Sample space of travel time from A to C = {8, 9, 10, 11, 12, 13, 14}
(b) Sample space of travel cost from A to C = {850, 1500}
(c) Sample space of TAC and CAC
= {(8, 1500), (9, 1500), (10, 1500), (11, 850), (12, 850), (13, 850), (14, 850)}

2.2
(a) Since the possible values of settlement for Pier 1 overlap partially with those of Pier 2, it is
possible that both Piers will have the same settlement. Hence, the minimum differential
settlement is zero.
The maximum differential settlement will happen when the settlement of Pier 2 is 10 cm and
that of Pier 1 is 2 cm, which yields a differential settlement of 8 cm.
Hence, the sample space of the differential settlement is zero to 8 cm.
(b) If the differential settlement is assumed to be equally likely between 0 and 8 cm, the
probability that it will be between 3 and 5 cm is equal to

P!

5"3 2
! ! 0.25
8"0 8

2.3
(a)

(b)
Wind
direction

E1
E2

90o

E3

60o

E1

30o
0

15

35

45

(c)

A and B are not mutually exclusive
A and C are not mutually exclusive

Wind Speed

2.4

(a) Inflow
6’
6’
6’
7’
7’
7’
8’
8’
8’

Possible water level
Outflow Inflow – Outflow + 7’
5’
8’
6’
7’
7’
6’
5’
9’
6’
8’
7’
7’
5’
10’
6’
9’
7’
8’

Hence possible combinations of inflow and outflow are
(6’,5’), (6’,6’), (6’,7’), (7’,5’), (7’,6’), (7’,7’), (8’,5’), (8’,6’) and (8’,7’).
(b) The possible water levels in the tank are 6’, 7’, 8’, 9’ and 10’.
(c) Let E = at least 9 ft of water remains in the tank at the end of day.
(8’, 5’) and (8’, 6’) are favourable to the event E.
So P(E) =

3 1
= .
9
3

Sample points (7’, 5’),

2.5
(a)
Locations Locations
MA
Probability
of W2
of W1

0
0
0
0.15x0.2=0.03

B
500
0
5,000
0.045

C
0
500
10,000
0.075
B

200
0
2,000
0.05
B
B
700
0
7,000
0.075
B
C
200
500
12,000
0.125
C

0
200
4,000
0.12
C
B
500
200
9,000
0.18
C
C
0
700
14,000
0.30

E1

X
X
X
X
X

E2

E3

X
X
X
X

X
X

X
X

(b) E1 and E2 are not mutually exclusive because these two events can occur together, for
example, when the weight W2 is applied at C, MA is 10,000 ft-lb; hence both E1 and E2 will
occur.
(c) The probability of each possible value of MA is tabulated in the last column of the table
above.
(d) P(E1) = P(MA > 5,000) = 0.075 + 0.075 + 0.125 + 0.18 + 0.30 = 0.755
P(E2) = P(1,000 ! MA ! 12,000) = 0.045 + 0.075 + 0.05 + 0.075 + 0.12 + 0.18 = 0.545
P(E3) = 0.05 + 0.075 = 0.125
P(E1 " E2) = P(5,000 < MA ! 12,000) = 0.075 + 0.075 + 0.18 = 0.33
P(E1 # E2) = 1 – 0.03 = 0.97
P( E 2 ) = 0.03 + 0.125 + 0.3 = 0.455
P( P ( E2 ) \$ 1 % P ( E ) \$ 1 % 0.545 \$ 0.455

15 P(B2!B1) = 0.05 + 0.0925)(0.2325 – (0.45)(0.P(A2!A1) P(A1) = 0.15 + 0.05 P(B1) = P(B2) = 0.P(B2!B1) P(B1) = 0.2.(0.P(A)P(B) = 0.6 (a) Let A1 = Lane 1 in Route A requires major surfacing A2 = Lane 2 in Route A requires major surfacing B1 = Lane 1 in Route B requires major surfacing B2 = Lane 2 in Route B requires major surfacing P(A1) = P(A2) = 0.05 .15) = 0.(0.15 P(A2!A1) = 0.0925 + 0.15 .05) = 0.15)(0.0925 P(Route B will require major surfacing) = P(B) = P(B1"B2) = P(B1) + P(B2) .45 P(Route A will require major surfacing) = P(A) = P(A1"A2) = P(A1) + P(A2) .2325 (b) P(route between cities 1 and 3 will require major resurfacing) = P(A"B) = P(A) + P(B) .302 .2325) = 0.

1x0.1 + 0.1 + 0.7 P(Di) = 0.1x0.9 = 0.9x0.9x0. the probability become P = 0.9 x 0.1 Assume condition between welds are statistically independent (a) P( D1 D 2 D3 ) = P( D1 )P( D 2 )P( D3 ) = 0.1x0.1)3 = 0.2.001 .729 (b) P(Exactly two of the three welds are defective) = P( D1 D2D3 ! D1 D 2 D3 ! D1 D2 D3 ) = P( D1 D2D3) + P(D1 D 2 D3) + P(D1 D2 D3 ) since the three events are mutually exclusive.1x0. Hence.9 = 0.027 (c) P(all 3 welds defective) = P(D1 D2 D3) = P3(D) = (0.9 x 0.

6.8 x 0.296 (c) P(casting !concrete production not feasible at site) = P((E2 " E3) E1! E 2 ) = P( E 2 " E 3 ) E1 E 2 P( E 2 ) = 0.2x(0.4x0.8 P(E1) = 0. assume E2 and E3 are statistically independent of E1 (a) A = (E2 " E3) E1 B = ( E 2 " E 3 ) E1 or E 2 E 3 " E1 (b) P(B) = P( E1 " E 2 E 3 ) = P( E1 ) + P( E 2 E 3 ) – P( E1 E 2 E 3 ) = 0.2 + 0.3) = 0.8. P(E2) = 0.3 – 0.2.6 =0.95 P(E3! E 2 ) = 0. P(E3) = 0.7.4x0.48 # P( E1 E 2 E 3 ) P( E 2 ) # P( E1 ) P( E 3 E 2 ) P( E 2 ) P( E 2 ) .2 + P( E 3 ! E 2 )P( E 2 ) – P( E1 )P( E 2 E 3 ) = 0.

1 is in good condition = E1 E 2 E 3 B = exactly one tractor is in good condition = E1 E 2 E 3 ! E1 E2 E 3 ! E1 E 2 E3 C = at least one tractor is in good condition = E1 ! E 2 ! E 3 (b) Given P(E1) = P(E2) = P(E3) = 0.0. E2. also the probability of each of these three events is the same.8 P(A) = P(E1 E 2 E 3 ) = P(E1" E 2 E 3 ) P( E 2 E 3 ) = [1 – P( E1 " E 2 E 3 )] P( E 2 " E 3 ) P( E 3 ) = (1 .144 P(C) = P(E1 ! E2 ! E3) = 1 – P( E1 ! E 2 ! E 3 ) = 1 – P( E1 E 2 E 3 ) = 1 – P( E1 " E 2 E 3 ) P( E 2 " E 3 ) P( E 3 ) = 1 – 0.6)(0.6 or P( E 2 " E 3 ) = 0. 3 are in good condition respectively (a) A = only tractor no.6 P( E 2 " E1 ) = 0. E1 E2 E 3 and E1 E 2 E3 are mutually exclusive.8)(0.8x0. 2.6x0. E3 denote events tractor no.048 = 0.4) = 0.048 Since E1 E 2 E 3 . 1.4 = 0. P(B) = 3 x P(E1 E 2 E 3 ) = 3 x 0.8 or P( E1 " E 2 E 3 ) = 0.6 P( E 3 " E1 E 2 ) = 0.808 .2.9 E1.

e.4 = 0.0.0. (iii) Given: P(A!B) = 0. i. However. (A’s being available boosts the chances that B will be available) (ii) From (a). P(AB) is nonzero.5)/(1 .e.5 (b) P(B is available#A is not available) = P(B# A ) = P( BA ) P( A ) while it is clear from the following Venn diagram that P( BA ) = P(B) . hence since P(A!B) = P(A) + P(B) .P(A!B) = 0.3/0. hence AB % &.6) = 0. using Bayes’ rule. we must have P(B#A) = P(B) = 0.e..9 " A!B does not generate the whole sample space (otherwise the probability would be 1).8333 So A and B are not s. A and B are not collectively exhaustive.6 + 0.8 .P(AB).10 (a) The event “both subcontractors will be available” = AB. i.9 = 0.5/0.8 .2. P(B#A) = P(AB)/P(A) = 0.6 = 0.75 (c) (i) If A and B are s. A and B are not m.i.i.P(AB) " P(AB) = P(A) + P(B) . .8.0. A AB B A Hence P( BA ) P( B) \$ P( AB) = 1 \$ P( A) P( A ) = (0.

SA.2.2#0.P(SA SB)] = P(L) [P(SA) + P(SB) .e.00046 . P(SB!SA) = 0. P(L!SB) = P(L). SB denote the respective events “leakage at site”. whose probability is P(L SA) + P(L SB) .01.P(L SA L SB) = P(L)P(SA) + P(L)P(SB) . P(SB!L) = P(SB). P(SB) = 0. i. whose probability is P(L"SA) = P(L! SA) P(SA) = P(L)P(SA) = 0.0002.P(L SA SB) = P(L) [P(SA) + P(SB) .03.02.P(SB!SA) P(SA)] = 0.02) = 0.01 (0. “seam of sand from X to A”.02 + 0.03 .11 (a) Let L. “seam of sand from X to B”.02 = 0. P(SA) = 0.01#0. Also given: independence between leakage and seams of sand. Given probabilities: P(L) = 0. P(L !SA) = P(L) The event “water in town A will be contaminated” = L"SA.2. (b) The desired event is (L SA) \$ (L SB).0. P(SA!L) = P(SA).

.2 (c) The event of interest is A#B#C. Since this is the union of many items.9.2. P( AB | C ) = 0. P(C) = 0.9!(1 – 0. allowing us to apply De Morgan’s rule and rewrite as P(A#B#C) = 1 – P( A # B # C ) = 1 – P( ABC ) = 1 – P( AB | C )P( C ) = 1 – 0.C denote the respective events that the named towns are flooded.12 Let A.6. Given probabilities: P(A) = 0. (a) P(disaster year) = P(ABC) = P(A | BC)P(BC) = P(A | BC)P(B | C)P(C) = 0.B.1 = 0.048 (b) P(C | B) = P(BC) / P(B) = P(B | C)P(C) / P(B) = 0.2.6!0.1"0.8!0.81 = 0.6!0.3 = 0. P(B) = 0. P(B | C) = 0.1) = 1 – 0.1.8. where an overbar denotes compliment of an event. we can work with its compliment instead.3. P(A | BC) = 0.19 by De Morgan’s rule.

P(C) = P[(L!M) G]] + P[(L"M) G ] = P(LG!MG) + P(L)P(M G ) = P(LG) + P(MG) – P(LMG) + P(L)P(M| G )P( G ) = P(L)P(G) + P(M|G)P(G) – P(L)P(M|G)P(G)+ P(L)P(M| G )P( G ) = 0.7#0.6 / 0.3#1#0.7#0.2.7#1#0. Hence P(C) is simply the sum of two terms.6 + 1#0.6 + 0.74 (c) P( L |C) = P( L C)/P(C) = P( L {[(L!M) G]} ![(LM) G ])) / P(C) = P[ L (L!M) G]] / P(C) impossible event = P( L MG) / P(C) = P( L )P(MG) / P(C) = P( L )P(M|G)P(G) / P(C) = 0.5#0.13 [(L!M) G]] ![(LM) G ] (a) C= (b) Since (L!M)G is contained in G.4 = 0.74 \$ 0.243 since ( L L)M G is an . it is mutually exclusive to (LM) G which is contained in G .6 – 0.

and 80% of (60 + 20) = 0. so D and R are not statistically independent.685 (d) (i) (ii) D and R are not mutually exclusive. but here P(R | D) = 30 / 90 = 1/3. There were a total of 30 + 20 + 60 + 20 = 130 accidents in 10 years. hence the total is P(F) = (25 + 64) / 130 " 0. hence P(S | D) = 60/90 = 2/3 (c) Let F denote “fatal accident”. no more than one accident per year.013 (probability per year) 1000 (b) Examining the data across the “Day” row.2. . we must have P(R | D) = P(R).5!50 = 25 fatal “run into train” accidents. we see that the relative likelihood of R compared to S is 30:60.14 (a) Note that the question assumes an accident either occurs or does not occur at a given crossing each year. while P(R) = (30 + 20) / 130 = 5/13. hence the yearly average is 130 / 10 = 13 accidents among 1000 crossings.8!80 = 64 fatal “struck by train” accidents. i. they can occur together (there were 30 run-into-train accidents happened in daytime).e. hence the yearly probability of accident occurring at a given crossing is 13 = 0. We have 50% of (30 + 20) = 0. If D and R are.

15 F = fuel cell technology successfully marketable S = solar power technology successfully marketable F and S are statistically independent Given P(F) = 0. P(S) = 0.7)(0.7.7 + 0.85 = 0.85) = 0.7x0.2.85 (i) P(energy supplied) = P(F!S) = P(F) + P(S) – P(F)P(S) = 0.85 – 0.85) + (1-0.955 (ii) P(only one source of energy available) = P(F S ! F S) = P(F S ) + P( F S) = (0.7)(1-0.36 .

0. P(at least one rainy day) = P(E1 " E2 " E3) = P(E1) + P(E2) + P(E3) .5x0.2 =0.52 Where P(E1E3)=P(E3|E1)P(E1)=0.375 .375x0.1125 P(E3|E1)= P(E3|E2E1)P(E2|E1)+ P(E3| E 2 E1)P( E 2 |E1) =0.2.1125 + 0.5x0.3) = 0. P(E1E2 E 3 ) = P( E 3 !E1E2) P(E2! E1) P(E1) = (1-0. P(E1E2) = P(E2! E1) P(E1) = 0.P(E1E3) + P(E1E2E3) = 0.15 c.12 d.3 P(E2! E1) = P(E3! E2) = 0.P(E2E3) .0.5)(0.5 .3x0.5 P(E3! E1E2) = 0.16 a.2 b.3 = 0.5 =0.P(E1E2) .5 – 0.2)(0.3x0.3x0.15x0. E1 = Monday is a rainy day E2 = Tuesday is a rainy day E3 = Wednesday is a rainy day Given P(E1) = P(E2) = P(E3) = 0.3 + 0.3 .15+0.3 + 0.3x0.3=0.

4 (b) P(able to park) = 1 – P(not able to park) = 1 – P( A B C ) = 1 – P( A B )xP( C ! A B ) = 1 – 0.789 0. B and C are available on a week day morning respectively Given: P(A) = 0. P(C) = 0.15.4 = 0.4x(1-0.8 P(B! A ) = 0. B.76 (c) P(free parking!able to park) P(A"B!A"B"C) = P[( A " B )( A " B " C )] P( A " B " C ) = P( A " B) P( A " B " C ) = 1 # P( AB) 0.4) = 1 – 0.2.8 x 0. P(C! A B ) = 0.17 Let A.76 . C denote events parking lot A. P(B) = 0.5 = 0.76 = 1 # 0.24 =0.4 (a) P(no free parking) = P( A B ) = P( A ) x P( B ! A ) = 0.2.5.

02 =0. P(C) = 0.11 .2.2x0.1 = 0.1 / 0.13 (b) P( EC!E " C) = P(( EC ) # ( E " C )) P( E " C ) = P(EC) / 0.05 P(C!E) = 0.05 + 0.2x0.13 = 0.13 – 0.154 (c) P(E C " E C) = P(E C ) + P( E C) = P(E"C) – P(EC) = 0.1.13 = 0.1 – 0.2 (a) P(Failure) = P(C " E) = P(C) + P(E) – P(C!E)P(E) = 0.18 C = Collapse of superstructure E = Excessive settlement Given: P(E) = 0.

05-0.[0.{1-[P(M)+P(W)-P(MW)]}P( B ) = {1 .5 = 0.02 – 0. P(B) = 0.{1 .P( M W B ) = 1.02+0. P(W) = 0.53 (c) P(only one component failed) = P( W M B) + P( W M B ) + P(W M B ) P(W M B ) = 0.05.P(M ! W)}P( B ) = 1.02.02 (b) P(system failure) = 1 – P(no component failure) = 1.5) = 0.5) = 0.01 B is statistically independent of M or W (a) P(W M B ) = P(W M )P( B ) = [P(W) – P(MW)] [1 – P(B)] = (0.2.19 M = failure of master cylinder W = failure of wheel cylinders B = failure of brake pads Given: P(M) = 0.005 also P( W M B) = P( W M ) x P( B ) = [1-P(M)-P(W)+P(MW)]P(B) .01)(1 – 0.02 from part (a) also P( W M B ) = P( W M) x P( B ) = [P(M)-P(MW)]P( B ) = (0.01]}(0.5 P(MW) = 0.P( M W ) P( B ) = 1.05 – 0.01)x0.

005+0.02-0.53 = 0.5 = 0.02) / 0.005+0.934 .01)x0. W M B and W M B are all within the event of system failure P(only one component failure!system failure) = (0.47 Since the three events W M B.= (1-0.47+0.

1 P(E2 !E1) = 0.4x0.933 .4x0.4 = P(E3 !E2) P(E3 !E1 E2) = 0.008 (d) P( E2 ! E1 ) = ? Since P( E1 " E2 ) = P( E1 )+P( E2 )-P( E1 E2 ) = P( E1 )+P( E2 )-P( E2 ! E1 )P( E1 ) = 0.8-0.96 and P( E2 ! E1 ) = 0.2 (b) P(E1 " E2) = P(E1)+P(E2)-P(E1 E2) = 0. 1.4 = 0.9-P( E2 ! E1 ) 0.1x0.9+0.1x0.96 Therefore.16 (c) P(E1E2E3) = P(E1) P(E2 !E1) P(E3 !E1 E2) = 0. P( E1 " E2 ) = 1.1 = 0.9 P( E2 ! E1 ) = 0.20 E1 = Excessive snowfall in first winter E2 = Excessive snowfall in second winter E3 = Excessive snowfall in third winter (a) P(E1) = P(E2) = P(E3) = 0.1 – 0.P(E1E2) = 1.2.2 =0.1 + 0.9 and from given relationship.P(E1) P(E2 !E1) = 1-0.

2x0.8 .P( H 2 ! H 1 )(0.85x0.21 H1.4 = 0.20.8x0.2.P(H2 !H1) P(H1) we have.4x0.032 (b) P( H 2 ! H 1 ) = ? Using the hint given in part (d) of P2. H2.85 (c) P(at least 1 hot summer) = 1 – P(no hot summers) = 1 – P( H 1 H 2 H 3 ) = 1 – P( H 1 )P( H 2 ! H 1 ) P( H 3 ! H 2 ) = 1 – 0.422 . H3 denote first. second and third summer is hot respectively Given: P(H1) = P(H2) = P(H3) = 0.8) = 1 – 0.2 therefore.4x0. P( H 1 ) + P( H 2 ) .2 P(H2 !H1) = P(H3 !E2) = 0. 0.85 = 0. P( H 2 ! H 1 ) = 0.8 + 0.4 (a) P(H1H2H3) = P(H1) P(H2 !H1) P(H3 !E2) = 0.P( H 2 ! H 1 )P( H 1 ) = P( H 1 " H 2 ) = 1 – P(H1H2) = 1 .

9[1-{0.05 (b) P(no power) = P(ABC) = P(AB)P(C) = P(B!A)P(A)P(C) = 0.1675 .5 = P(A!B) C is statistically independent of A and B (a) P(complete backout!A) = P(BC!A) = P(B!A)P(C) = 0. P(B) = 0.05-0.1 P(B!A) = 0.05.1 = 0.22 A = Shut down of Plant A B = Shut down of Plant B C = Shut down of Plant C Given: P(A) = 0.1 =0.5x0.05x0.05.05}] = 0.P(B!A)P(A)}] = 1 – 0.5x0.2.5x0.00025 (c) P(less than or equal to 100 MW capacity) = P (at most two plants operating) = 1 – P(all plants operating) = 1 – P( A B C ) = 1 – P( C )P( A B ) = 1 – P( C ) [1-{P(A)+P(B).05+0. P(C) = 0.

98x0.2.02 = 0.23 P(Damage) = P(D) = 0.023 = 8x10-6 (b) P( D1 D2) = P( D1 )P(D2) = 0.02 Assume damages between earth quakes are statistically independent (a) P(no damage in all three earthquakes) = P3(D) = 0.0196 .

8"0.98 \$ 0.2.75)3 \$ 0.8"0. hence P(error spotted) = 1 – (1 – 0.8 + 0. the probability of not finding the error is P(C1’C2’C3’) = P(C1’)P(C2’)P(C3’) = (1 – 0. so the team of 3 is better. P(E) = P(A!B) = P(A) + P(B) – P(AB) = P(A) + P(B) – P(A)P(B) = 0. Hence P(D1 D2) = P(D1)P(D2) = 0.960 .75)3 .1 / 0. Let E denote the event that the error is spotted. detection of the second error (event D2) 2 still has the same probability.9 = 0.98 (b) “Spotted by A alone” implies that B failed to spot it.98 in (a). B denote the event of the respective engineers spotting the error. which is higher than the 0.9 – 0.98. However.082 (c) With these 3 engineers checking it. hence the required probability is P(AB’|E) = P(AB’#E)/P(E) = P(E|AB’)P(AB’) / P(E) = 1"P(A)P(B’) / P(E) = 0.24 (a) Let A. (d) The probability that the first error is detected (event D1) has been calculated in (a) to be 0.984. since statistical independence is given.98 \$ 0.

406 (b) P(only one of the two potential failure modes) = P(A L ) + P( A L) = P( L !A)P(A) + P( A !L)P(L)(0.2.006 P(L!A) = 0.0148 P( A " L) P ( A " L) 0.006x0.4)(0.008) = (1-0.4 P(A!L) = 0.25 L = Failure of lattice structure A = Failure of anchorage Given: P(A) = 0.0.0092 (c) P ( A D) # P( A( A " L)) P( A) 0. P(L) = P(L!A)P(A) / P(A!L) = 0.008 (a) P(antenna disk damage) = P(A"L) = P(A) + P(L) – P(A)P(L!A) = 0.4 + 0.3)(0.406 .0036 + 0.0056 =0.006 # # # 0.006) + (1-0.008 .008) = 0.3 Hence.4 = 0.

26 P(F) = 0.1 P(A!F) = 1 (a) FA.99 = 0.9x0.01 = 0 P( F A) = P(A! F )P( F ) = 0.99 = 0.01 " 0.01 + 0.0917 0.01 P(F A ) = P( A !F)P(F) = 0x0.099 P( F A ) = P( A ! F )P( F ) = 0. F A.01 P(A! F ) = 0.99 = 0. F A . F A are set of mutually exclusive and collectively exhaustive events (b) P(FA) = P(A!F)P(F) = 1x0.891 (c) P(A) = P(A!F)P(F) + P(A! F )P( F ) = 1x0.109 .109 (d) P(F!A) = P( A F ) P( F ) P( A) " 0.2.01 = 0.1x0.1x0.

01998 .27 Given: P(D) = 0.15x0.00085 # 0.0408 0.00085 P(D T ) = P( T !D)P(D) = 0.001 = 0.001 P(T!D) = 0.00015 P( D T) = P(T! D )P( D ) = 0. D T (b) P(DT) = P(T!D)P(D) = 0.001 = 0.979 (c) P( D T ) " P(T D) P( D) P(T D) P( D) # P(T D) P( D) " 0.98x0.85x0.999 = 0.01998 P( D T ) = P( T ! D )P( D ) = 0.85.999 = 0. P(T! D ) = 0. D T.2.02x0.00085 " 0. D T .02 (a) DT.

P(RA) P(RB!RA) From the hint given in P2.5 P(RB) = P(GB) = 0.9 .5] \$ 1 # 0.20.2 Signal at C is statistically independent of those at A or B.5x0.2x0.2.8 % 0.8 = 0.8 P(GLT) = 0.28 Given: P(RA) = P(GA) = 0.5 Hence P(stopped at most once from M to N) = 1 – 0.92 (c) P(stopped at most once from M to N) = 1 – P(stopped at both A and B) = 1 – P(RARB) = 1 .05 P(GB!GA) = 0.5x0. (a) E1 = RA" RB E2 = GLT E3 = RA GB " GA RB (b) P(stopped at least once from M to Q) = 1 – P(no stopping at all from M to Q) = 1 – P(GA GB GLT) = 1 – P(GLT)P(GA)P(GB !GA) = 1 – 0.2 P( R A ) 0 . it can be shown that P( RB R A ) # 1 1 [1 \$ P(G B G A ) P(G A )] \$ 1 # [1 \$ 0.2 = 0.

1!0.05x0.29 (a) P(WSC) = P(W)P(SC) = P(W)P(S|C)P(C) = 0.8!0.05 = 0.05 + 0.9 = 0.05 = 0.1 + 0.1 + (1 – 0.1!(1 – 0.05 = 0.5xP(W)P( C ) + 0.5x0.3)!0.0315 (e) P(U) = P(U |CW)P(CW) + P(U | C W)P( C W) + P(U|C W ) P(C| W ) + P(U|| C W ) P( C W ) = 1xP(W)P(C) + 0.5xP(C)P( W ) + 0xP( C )P( W ) = 1x0.05 – 0.5x0.05] / (1 – 0.1x0.0015 (b) P(W"C) = P(W) + P(C) – P(WC) = P(W) + P(C) – P(W)P(C) = 0.95 + 0.145 P(W"C| S ) = P[ S (W"C)] / P( S ) = P( S W" S C) / P( S ) = [ P( S W) + P( S C) – P( S WC) ] / P( S ) = [ P( S )P(W) + P( S |C)P(C) – P(W)P( S C) ] / P( S ) = [0.3)!0.7x0.05 – 0.2.075 .9x0.1x0.139 (c) (d) P(nice winter day) = P( S W C) = P( W )P( S C) = P( W )P( S | C)P(C) = 0.2) # 0.1!0.3!0.

662 .0592 # 0.2. P(C) = P(L!B) = P(L) + P(B) – P(LB) = P(L) + P(B) – P(L)P(B)! L and B are independent events = 0.0592 (b) P(L B | C) = P(CL B ) / P(C) = P(C | L B )P(L B ) / P(C).30 (a) Let L and B denote the respective events of lead and bacteria contamination.04"(1 – 0.04 + 0. but P(C | L B ) = 1 (! lead alone will contaminate for sure) and P(L B ) = P(L)P( B ) (! statistical independence).02) / 0.02 = 0.04"0. hence the probability = 1"P(L)P( B ) / P(C) = 1"0. and C denote water contamination.02 – 0.

5x0. 42.1 0. 20.4=0.5x0.2 0. 32.5x0.1 = 0.2 .2 0.2 Water level 0 10 20 0 10 20 W (x103 lb) 100 210 320 100 210 320 F (x103 lb) 10 21 32 20 42 64 (a) Sample space of F = {10.1x0.1 0.2.2 0.1 0.5 + 0.06 Probability 0.31 C 0.5x0.5x0.4=0.2 0.1 0. 21. 64} (b) Let H be the horizontal force in 103 lb P(sliding) = P(H>F) = P(F<15) = 0.2=0.5x0.1 (c) P(sliding) = P(F<15) P(H=15) + P(F<20) P(H=20) = 0.2 0.4=0.4=0.2=0.1 0.1x0.

the union of I and II does not make up the entire sample space.056 P( E W H) = P( E ) P( W !H) P(H) = 0.3-0. I and II are mutually exclusive.3) = 0.4x0.32 Given: (a) P(W) = 0.3 P(E) = 0.096 (c1) Since P(W!H) = 0.6 # 0.3 = 0.4x0.6 # 0.9 = P(W). . W and H are not statistically independent (c2) I II = E( W "H) ( E W H) = (E E )( W "H)( W H) Since E E is an empty set. Since P(W!H) = 0.2 P(W!H) = 0.2x(0.9 P(H) = 0. From the above Venn Diagram. W and H are not mutually exclusive. I and II are not collectively exhaustive. Hence.6 E is statistically independent of W or H I = E( W "H) II = E W H (b) P[E( W "H)] = P(E) P( W "H) = P(E) x [P( W )+P(H)-P( W !H)P(H)] = 0.2.1+0.8x0.

056 + 0.096 = 0.152 .(d) Since I and II are mutually exclusive. P(leakage) = P(I) + P(II) = 0.

1x0.1x0.3x0.15 = 0.006/0.85 = 0.1 P(O) = 0.2 = 0.33 Given: P(E) = 0.2x0.2 + 0.15 + 0.15 P(G) = 0.1x0.2 P(E!O) = 2x0.083 .235 (c) P(GO!E) = P(GOE)/P(E) = 0.006 = 0.9x0.006 (b) P(G " E) = 1-P( G E ) = 1.P( G ) P( E ) = 1-0.15 = 0.2.04 (d) P(at least 2 sources will be in short supply) = P(EG"GO"EO) = P(EG)+P(GO)+P(EO)-P(EGO)-P(EGO)-P(EGO)+P(EGO) = P(G)P(E)+P(G)P(O)+P(E!O)P(O)-2P(EGO) = 0.3x0.3 G is statistically independent of E or O (a) P(shortage of all three sources) = P(EGO) = P(G) P(E!O) P(O) = 0.2 .

P(BC!H) .16308x0. B. B.03 – 2x0. P(S!H) = P(A!H) + P(B!H) + P(C!H) .34 A.02 + 0.P(AB!N) .03x0.096636 Similarly.07.08x0.2. C denote failure of component A.03x0. P(C!H) = 0. P(B!H) = 0.08x0.03.05x0.08x0.08 + 0.6 \$ 0.02 – 2x0.02x2x0.03 & 0.08 = 0.03 + 0.08.02 P(A!H) = 0.08 & 0.172 0. C respectively N and H denote normal and ultra-high altitude respectively Given: P(A!N) = 0.03 – 0.02 – 0.123 P( B S ) % = P[ B( A " B " C )] P( B ) P ( B N ) P( N ) \$ P( B H ) P ( H ) % % P( S ) P( S ) P( S ) 0.03x2x0.03 = 0.96636x0.07x0.6 + 0.6.P(AC!H) + P(ABC!H) = 0.P(BC!N) .03 – 0.08 – 0.123 .16308 Hence P(system failure) = 0.05.4 P(B!A) = 2 x P(B) C is statistically independent of A or B P(system failure) = P(S) = P(S!N)P(N) + P(S!H)P(H) P(S!N) = P(A"B"C!N) = P(A!N) + P(B!N) + P(C!N) . P(B!N) = 0.03 + 0.03 P(N) = 0. P(C!N) = 0.P(AB!H) .05 + 0. P(H) = 0.4 % 0.03x0.4 # 0.P(AC!N) + P(ABC!N) = 0.07 + 0.

5 = 0.4 % 0.5 = 0.3 P(W!C) = 0.0.5 P(W) = 0.4 # 0.6 0.4 U=C"W (a) Since P(W!C) = 0.6 (c) P(C W ) = P( W !C) P(C) = (1-0.4x0.4)x0.2.333 0.6 P(C " W ) P (C " W ) .3=P(W) W and C are not statistically independent (b) P(U) = P(C " W) = P(C) + P(W) – P(W!C) P(C) = 0.35 Given: P(C) = 0.5 P[CW (C " W )] P(CW ) \$ \$ \$ \$ 0.5 + 0.3 (d) P (CW C " W ) \$ P (W C ) P(C ) 0.3 .

6.6x0.15x0.0.4 = 0.6 + 0.422 (c) If C is congested.3x(1-0.85 C is statistically independent of A or B P(L!ABC) = 0.6. P(exactly one route congested) = 0. C are congested respectively Given: P(A) = 0. P(ABC) = 0.3x0.85x0.51 P(L) = 0. C denote route A.204) = 0.606 . P(C) = 0.85x0. B. B.6x1 = 0.3 (a) P(A B C )+ P( A B C )+ P( A B C) = P( B !A)P(A)P( C )+P( A !B)P(B)P( C )+P( A B )P(C) but P( A B ) = 1-P(A "B) = 1-P(A)-P(B)+ P(B!A)P(A) = 1 .204 + 0.9x0.9x0.0.4 = 0. P(L) = 0.31x0.2.6 = 0.85x0.6 +0.6 .204 Hence. P(B) = 0.51 + 0.6x0.15x0.4 P(A!B) = P(B!A) = 0.6x0.36 A.6 + 0.31 Hence.232 (b) P(Late) = P(L!ABC)P(ABC)+ P(L! ABC ) P( ABC ) But P(ABC) = P(AB)P(C) = P(B!A)P(A)P(C) = 0.49 = 0.9 P(L! ABC ) = 0.

2 + 0.188 (b) The required probability is [P(H2 L1) + P(H2 N1)] / P(heavy damage after two quakes) = (0.05!0.143 . Alternatively.05!0.05!0. P(heavy damage after two quakes) = P(H2 H1) + P(H2 L1) + P(H2 N1) = P(H2 | H1) P(H1) + P(H2 | L1)P(L1)+ P(H2 | N1)P(N1) = 1!0.143.2 – 0. (ii) H. Hence P(any heavy damage after 3 quakes) = 1 – P(no heavy damage in each of 3 quakes) 3 = 1 – 0.75) / (1!0.05!0. one could explicitly sum the probabilities of the three mutually exclusive events.95 " 0. Note that (i) occurrence of H1 makes H2 a certain event.05 + 0.75) " 0.5!0.95!0.05 + 0.2 + 0.05) " 0.05!(1 – 0.05 + 0.95 (not influenced by previous condition).2.05) = 0. one always starts from an undamaged state.37 (a) Let subscripts 1 and 2 denote “after first earthquake” and “after second earthquake”.5!0.95 + 0.95 " 0. the probability of not getting heavy damage at any stage is simply (1 – 0.2 + 0. L and N are mutually exclusive events and their union gives the whole sample space.5!0. P(H) = P(H1) + P(H2 H1’) + P(H3 H2’ H1’) = 0.733 (c) In this case.

92x0.1 \$ 0.92 .1 = 0. P(H) = 0. P(S!A) = 0.92 = 0.38 Given: P(L) = 0.9.375 1 # 0. P(A) = 0.9x0. i.6.154 P( S A) P( A) P(S ) " 0.2. P(S!H) = 0.6 + 0.1 S denote supply adequate. no shortage P(S!L) = 1.92 (b) P( A S ) " (c) P(shortage in at least one or of the next two months) = 1 – P(no shortage in next two months) = 1 – 0.5x0.3 + 0.3.5 (a) P(S) = P(S!L)P(L)+P(S!A)P(A)+P(S!H)P(H) = 1x0.e.3 " 0.

5 + 0.2.3 = 0.39 (a) P(shipment accepted on a given day) = P(at most 1 defective panel) = 0.5 = 0.8 = 0.7 (b) P(exactly one shipment will be rejected in 5 days) = 5xP(0.36x0.36 (c) P(acceptance of shipment on a given day) = P(A) = P(A!D=0)P(D=0)+ P(A!D=1)P(D=1)+ P(A!D=2)P(D=2) in which D = number of defective panels on a given day P(A!D=0) = 1 P(A!D=1) = 1 P(A!D=2) = 1-P(both defective panels detected) = 1 – 0.7)4(0. P(A) = 0.36 Hence.2 + 0.3) = 0.2 + 0.8x0.808 .

5 = 0.40 The possible scenario of the working condition and respective probabilities are as follows: (a) P(completion on schedule) = P(E) = P(E!G) P(G) + P(E!N) P(N) + P(E!BC) P(BC) + P(E!B C ) P(B C ) = 1x0.4 / 0.2x0.2 + 0.9x0.4 + 0.5 + 0.76 = 0.474 .2.8x0.4x0.4x0.76 (b) P(Normal weather!completion) = P(N!E) = P(E!N) P(N) / P(E) = 0.9x0.

3 + 0. probability of energy crisis = P(E!N) = P(O"G) = P(O)+P(G)-P(OG) = P(O)+P(G)-P(G!O)P(O) = 0.2 = 0. also O and G denote oil and gas supply is low respectively. (a) Given normal energy demand.1 = 0. H denote the event of low. normal and high demand respectively.6 +1x0.5x0.5 (b) P(E) = P(E!L) P(L) + P(E!N) P(N) + P(E!H) P(H) = P(OG)x0.43 .3 + 0.41 Let L.2 + 0.2.4 – 0.5x0. N.1 = P(G!O)P(O)x0.3 + 0.

196 + 0. P(H=2) = 0.75 Where H = number of hurricanes in a year P(J=H=1) = P(D!H=1) = 0.2 + 0.75 + 0.75 " " 0.0.2.99 P(J=H=2) = P(D!H=2) = 0.8x0.99x0. H=1 P(damage) = 1-P(JD!H=1) = 1. P(H=0) = 0.99x0.42 Given: P(H=1) = 0.80 Where J and D denote survival of jacket and deck substructure respectively Assume J and D are statistically independent (a) For the case of one hurricane.978 .75 + 0.99x0.99 = 0.8x0. P(no damage) = P(JD!H=0)P(H=0)+ P(JD!H=1)P(H=1)+ P(JD!H=2)P(H=2) = 1x0.P(J!H=1) P(D!H=1) = 1 .032 = 0.0199 (b) For next year where the number of hurricanes is not known.2.978 (c) P(H=0!JD) = P( JD H " 0) P( H " 0) 1 # 0.e.767 P ( JH ) 0.05 = 0. i.05.

HA.428 . B respectively (b) Assume the events of F.6. P(C! EA)=0 P(C! FB)=0.43 (a) Sample space = {FA. EB denote empty truck in lane A. FB denote fully loaded truck in lane A.1. 3 6 18 16 5 1 P( H A ) " . EA. P(C) = 1x5/18 + 0. P( H B ) " . P(C! HA)=0. B respectively HA. HB. B respectively EA. P(C! EB)=0 1 5 5 # " 3 6 18 1 P( E B ) " 16 P(FA) = P(fully loaded!truck in lane A) P(truck in lane A) = 1 1 1 1 # " .6x1/18 + 0. HB denote half loaded truck in lane A. P( E A ) " 18 18 Similarly P(FB) = Hence. H and E are equally likely P(full!truck in lane B) = 1/3 (c) Let C denote event of critical stress P(C) = P(C! FA)P(FA)+ P(C! HA)P(HA)+ P(C! EA)P(EA) +P(C! FB)P(FB)+ P(C! HB)P(HB)+ P(C! EB)P(EB) But P(C! FA)=1. FB.2. EB} in which FA.1x1/18 + 0 =0.4x5/18 + 0 + 0. P(C! HB)=0.4.

135 Since the failure in the shear mode was only 29.95 = 0.296 0.05 + 0.8 P(D!B) = 0.44 S denotes shear failure B denotes bending failure D denotes diagonal cracks Given: P(D!S) = 0.2.135 (b) P( S D) " P( D S ) P( S ) P( D) " 0. it does not exceed the threshold value of 75% required.8x0.05 " 0. .6%.1x0.8 # 0.1 5% of failure are in shear mode 95% of failure are in bending mode (a) P(D) = P(D!S)P(S) + P(D!B)P(B) = 0. immediate repair is not recommended.

20.80 + 0. P(R) = P(R | A)P(A) + P(R | D)P(D) + P(R | I)P(I) = 0.2.20 = 0.e.e.20!0.95)!0.048 (d) 100000 vehicles ! 5% = 5000 vehicles are expected to encounter the yellow light annually.60 and P(V’) = 0. i.001 = 0.05!0.048 = 1. and let C denote “Cautious driver in the other vehicle”. 5000!0.45 (a) Let A.006 (b) The desired probability is P(A | R). Let R denote the event that s/he will run the red light. also the conditional probabilities P(R | A) = 0.006 = 0.006 " 0. which can be found by Bayes’ Theorem as P(A | R) = P(R | A)P(A) / P(R) = 0.80 and P(C’) = 0.048 chance of getting into a collision (i. accident).08.005 + 0.02!0.80 + (1 – 0. 0.60 = 0.85. I denote the respective events that a driver encountering the amber light will accelerate. D. Out of these 5000 vehicles. The given probabilities are P(A) = 0.10.006 = 30 vehicles.05.05!0.10 / 0. where P(V) = 0.05.02 By the theorem of total probability.40.05!0.20 = 0.80)!0. hence 30!0. decelerate. hence P(collision) = P(collision | V)P(V) = 0.006) are expected to run a red light.6% (i.005 / 0. or be indecisive. P(I) = 0. The probability of collision is P(collision) = P(collision | V)P(V) + P(collision | V’)P(V’) But the second term is zero (since there is no other vehicle to collide with).08!0.e. 0.05 = 0. where P(C) = 0.44 accidents caused by dangerous vehicles can be expected at the intersection per year . These 30 dangerous vehicles have 0. while P(collision | V) in the first term is P(collision | C)P(C) + P(collision | C’)P(C’) = (1 – 0.10 + 0 + 0. P(R | D) = 0. P(R | I) = 0. P(D) = 0.833 (c) Let V denote “there exists vehicle waiting on the other street”.

B.628 (c) Given A is not profitable.1398 + 0.284 .7x0.336 (b) P(at least two branches profitable) = P(T) = P(ABC) + 0. B.8x0.714 P(bonus received) = P(E) = P(E!T)P(T)+P(E! T )P( T ) = 0.6 = 0. C will be profitable respectively E denotes bonus received P(A) = 0.8602 = 0.4 + 0.9x0.1x0.336 = 0.2.6 = 0.1 # 0.1x0.7x0.1398 P(bonus received) = 0.286 = 0.9x0.9 = P(B!A) C is statistically independent of A or B (a) P(Exactly two branches profitable) = P(AB C )+P(A B C)+P( A BC) = P(A!B)P(B)P( C )+P( B !A)P(A)P(C)+P( A !B)P(B)P(C) = 0.233 0.3 Hence P(T! A ) = 0.336 = 0.2x0. P(B) = 0.7x0. P(T! A ) = P(BC! A ) = P(B! A )P(C) But P ( B A) " P( A B) P( B) P( A) " 0.336 = P(A!B)P(B)P(C) + 0.2x0.7. C denote branch A.7x0.714 + 0.7 " 0.7. P(C) = 0.6 + 0.46 A.8x0.6 P(A!B) = 0.233x0.6 + 0.

5x0.5 P(C) = P(project in Champaign County) = 1 – 0.333 P(I) = P(project in Iroquois County) = 2/5 =0.333 = 0.4 " P( D !C)x0.167 # P( D !C) = 0.4 \$ 0.167 = 0.267 = 0.2.4 = 0.5x0.267 + 0.104 (c) P( I D) \$ P( D I ) P( I ) P( D) \$ 0 % 0.333 – 0.4 P(D!I) = 1.0 P(D!F) = 0.625x0.104 .0 0.333 + 0x0.47 P(D) = P(difficult foundation problem) = 2/3 P(F) = P(project in Ford County) = 1/3 = 0.267 (a) P(F D ) = P( D !F)P(F) = 0.625 Hence P(C D ) = 0.333 = P( D !C)x0.167 (b) P(C D ) = P( D !C)P(C) But P( D ) = P( D !C)P(C)+ P( D !F)P(F)+ P( D !I)P(I) 0.

0044 .1x0.0015 = 1.5x10-6 (b) P(E"L"T) = 1-P( E L T ) = 1-P( T )P( E " L ) = 1 – 0.001x0.001 # # # 0.227 P( E " L " T ) 0.9985x[1-P(E)-P(L)+P(L!E)P(E)] = 1 – 0.0044 (c) P( E E " L " T ) # P ( E ( E " L " T )) P( E ) 0.9971 = 0.48 P(E) = 0.002.002+0.0015 P(L!E) = 0.9985x0.001.1 T is statistically independent of E or L (a) P(ELT) = P(L!E)P(E)P(T) = 0.001-0.0015[1-0. P(T) = 0.001] = 1 – 0.1x0.2. P(L) = 0.0044 0.

1!0. E2 denote the respective events of using 100 and 200 units.12 = 0. P(S) = P(S | E1)P(E1) + P(S | E2)P(E2) = 0.49 (a) Let E1.06 / 0. and S denote shortage of material.4 = 0.2. P(E1 | S) = P(S | E1)P(E1) / P(S) = 0.18 = 1/3 .3!0.6 + 0.18 (b) Using Bayes’ theorem with the result from part (a).06 + 0.

2 (i) The “updated” probability of hard ground. but since ground is either hard or soft ! P(H) = 0.4 (ii) Using the updated probabilities P’(H) = 0. P(L | S) = 0. P(H) = 3P(S) .2.0.2"0.8 (b) Let E denote “penetration”.5. P’(H) # P(H | E) = P(E | H)P(H) / [P(E | H)P(H) + P(E | S)P(S)] by Bayes’ theorem = 0. the updated probability of a successful landing now becomes P’(L) = P(L | H)P’(H) + P(L | S)P’(S) = 0. P(L) = P(L | H)P(H) + P(L | S)P(S) = 0.25 (a) Using theorem of total probability.50 Let H and S denote Hard and Soft ground.9.9"0. and let L denote a successful landing.4.4 = 0. Given: P(E | S) = 0. P’(S) = 1 .75 + 0.9"0.5"0.75 / (0.6 = 0.75 + 0.9"0.6. P(S) = 0. Given probabilities: P(L | H) = 0.5"0.4 + 0.9. P(E | H) = 0.25 = 0.25) = 0.66 . respectively.2"0.75.

05 Hence P(S"C) = 0.05 + 0.05 = 0.1225 .P(C!S)P(S) = 0.05 Where C.51 P(C) = 0.1225 .45 0.2.095 (c) P(S S " C ) # P( S ( S " C )) P(S ) 0.P(C!S)P(S) But P( C !S) = P( S C ) P(C ) P(S ) # 0.1225 (b) P(C S " C S) = P(C S )+P( C S) = P(C"S) – P(CS) from Venn diagram = 0.025 \$ 0.5x0.1.9 # 0.1 – 0.408 P(S " C ) P( S " C ) 0.55x0.05 = 0.025 (a) P(S"C) = P(S)+P(C).55x0.05 # # # 0. P(S) = 0.1225 – 0.05 = 0. S denote shortage of cement and steel bars respectively P(S! C ) = 0.

2.3 + 0.3 \$ 0.3 # 0.36 (ii) P(A B " A B) = P(A"B) – P(AB) = P(A)+P(B)-2P(B!A)P(A) = 0.3 = 0.25.15 + 0 + 0.22 (iii) P(S) = P(S!A B )P(A B )+P(S! A B)P( A B)+P(S! A B )P( A B )+P(S!AB)P(AB) But P ( A B ) # P( B A) P( A) P( B) # 0. P(S! A B ) = 0.3x0.8 Where S denotes event of water shortage (i) P(A"B) = P(A)+P(B).06 P( A B ) = 1-P(AB)-P( A B)-P(A B ) = 1 – 0.P(B!A)P(A) = 0.3 \$ 0.15 .8x0.7x0.3 + 0.8 \$ 0.7x0.3 – 0.3x0.4x0.15 P(B!A) = 0.3 P(S!A B ) = 0.3.3 =0.0. P(B) = 0.129 (iv) P ( AB S ) # (v) P( S AB) P( AB) P( S ) # 0.129 P( A ! S ) = P( A B! S )+P( A B ! S ) = P( S AB ) P( AB) P( S ) % P( S AB) P( AB) P( S ) But P( A B) = [1-P(A!B)] P(B) = 0.6 0.06 – 0. P(S! A B) = 0.3x0.3x0. P(S!AB) = 0.25x0.52 Let A and B be water supply from source A and B are below normal respectively P(A) = 0.558 0.3 = 0.64 .3 # 0.15 –2x0.2.15 Hence P(S) = 0.15 = 0.3 + 0.4x0.3 = 0.2x0.

64 # ! 0.Hence P ( A S ) ! 0.871 .75 " 0.871 0.06 1 " 0.786 0.

66 (b) P(C) = P(C!T1 T2) P(T1 T2)+ P(C!T1 T2 ) P(T1 T2 )+ P(C! T1 T2) P( T1 T2)+ P(C! T1 T2 ) P( T1 T2 ) = 0.3x0.6 = 0. (a) P(C! T1 T2 ) = 0.9x0.9.4x0.3 = 0.729 (c) P(Deadline met) = P(FC) = P(F)P(C) = 0.7x0. P(F! W ) = 0.3 + 0.4 P(F) = P(F!W)P(W) + P(F! W )P( W ) = 0.6x0.6 P(C!T1 T2) = 0.9.4 + 0.5 P( W ) = 0.2.5x0.7 P(C!T1 T2 ) = P(C! T1 T2) = 0.53 Let W be weather favorable F be field work completed on schedule C be computation completed T1 be computer 1 available T2 be computer 2 available Given: P(F!W) = 0.6 P(T1) = P(T2) = 0.729 = 0.66x0.7 + 0.481 .7x0.7 + 0.9x0.3x0.6x0.

175 + 1x0.375 .3 = 0.) N = number of trucks in queue P(T<5!N=2) = P(loading time for both trucks ahead will take only 2 min.5 = 0.2.125 + 0.25x0. Hence P(T<5) = 1x0. the waiting time will definitely exceed 5. each) = 0.5x0. Hence those items will not contribute any probabilities.54 (a) T = wait time in queue (in min.25 (b) P(T<5) = P(T<5!N=0)P(N=0)+ P(T<5!N=1)P(N=1)+ P(T<5!N=2)P(N=2) Note that if the number of trucks in queue is 3 or more.

If one adopts the second alternative. The given probabilities are P(O1) = P(O2) = 0.06 + 0.06)!0.001!(1 – 0.001.05!(1 – 0.3.1)!(1 – 0.3!0.06!(1 – 0.1.05!0. P(D | O1’O2’) = 0.06!0. (a) P(D) can be determined by the theorem of total probability as P(D| O1O2)P(O1O2) + P(D| O1O2’)P(O1O2’) + P(D| O1’O2)P(O1’O2) + P(D| O1’O2’)P(O1’O2’) = 0.05!0.937 (c) The first alternative reduces all those conditional probabilities in (a) by half.06)!(1 – 0.1!0.1)!0.2.0128 / 2 = 0. hence the P(D) also reduces by half to 0.9!0. respectively.06) " 0.9 / 0.05!(1 – 0.06 + 0.001!0. and let D denote the event of bridge damage.1) " 0. P(D | O1O2) = 0.01281 " 0. P(D | O1O2’) = P(D | O1’O2) = 0.0128 (b) P(O1 # O2 | D) = 1 – P[ (O1 # O2 )’ | D] (but De Morgan’s rule says (O1 # O2 )’ = O1’O2’) = 1 – P(O1’O2’| D) = 1 – P(D | O1’O2’)P(O1’O2’) / P(D) = 1 – 0.55 Let O1 and O2 denote the events of truck 1 (in one lane) and 2 (in the other lane) being overloaded.05. . P(D) becomes P(D| O1O2)P(O1O2) + P(D| O1O2’)P(O1O2’) + P(D| O1’O2)P(O1’O2) + P(D| O1’O2’)P(O1’O2’) = 0.1!(1 – 0.3!0.1) + 0.06) + 0.0064.0076 Hence one should take the first alternative (strengthening the bridge) to minimize P(D).1 + 0.001!(1 – 0.1 + 0.

5 = 0.992 (iii) A failure probability of p means that out of a large number (N) of similar systems. We need P(A | D’) = P(D’ | A)P(A) / P(D’) by Bayes’ theorem.5"0.85.2.959 # 0.80"(1 – 0. in which case the expected loss would be 0. given no detection. the probability of having no anomaly. We are given P(D | A) = 0.85 (thus P*(A’) = 0. hence P(A | D’) = 0. p"N of them are expected to fail. The total failure cost would be p"N"\$1000000.0001 chance of failure.959 (ii) Total probability of a safe foundation = P(safe | A)P**(A) + P(safe | A’)P**(A’) = 0. it has only 1 – 0. and also P(D | A’) = 0 ! P(D’| A’) = 1.5 ! P(D’ | A) = 1 – 0.992)"\$1000000 # \$8300 If the system is known for sure to be anomaly free.85) + 0.85).9999"0.8 (thus P(D’ | A) = 0.3 + 1"0. however.0001"\$1000000 = \$100.7 = 0. Hence.85)] # 0. and also a better detection probability P(D | A) = 0.15/0.\$100) = \$8200 in expected loss is saved when the site can be verified to be anomaly free . (\$8300 .85) / [1"(0. in which P(D’) = P(D’ | A)P(A) + P(D’ | A’)P(A’) = 0.56 (a) Let A = “presence of anomaly” and D = “detection of anomaly by geophysical techniques”.85 # 0.3 / 0.7/0.9999 = 0.5.5"0.176 (b) (i) With the updated anomaly probability P*(A) = 0.2"(0. which when divided into N gives the “average” or expected cost per system = p"\$1000000. Hence we have the formula Expected loss = (probability of failure) " (failure loss) = (1 – 0. is now P**(A’) = P(A’ | D’) = P(D’ | A’)P*(A’) / P(D’) = 1"(0.7/0.7/0.959) + 0.2).15/0.

59 P’(B) = P(B!T) = 0.3 0.9 \$ 0.001x0.59 + 0.1 P(T!G ) P(G ) P(T!G ) P(G ) " P(T!A) P( A) " P (T!B) P( B) 0.2.71 = 0.71 = 0.1x0.2x0.001x0.57 Given: P(G) = 0.03 = 0.03 Hence P’(F) = 0.01 = 0.0003 + 0.38 0. P(A) = 0. P (F!A) = 0.01x0.001.0703 (b) T = passing the test P(T!G) = 0. P(B) = 0.9 \$ 0.7x0. (a) P(F) = P(F!G)P(G)+ P(F!A)P(A)+ P(F!B)P(B) = 0.38 + 0.3 " 0.01.2 (i) P(G!T) = = (ii) P(F!B) = 0.1 / 0.2 \$ 0.9.01x0.27 # # 0.71 P’(F) = P(F!G)P’(G)+ P(F!A)P’(A)+ P(F!B)P’(B) But P’(G) = P(G!T) = 0.7.3 + 0.7 \$ 0.3.06 +0.1 P(F!G) = 0.6 / 0.6 + 0.P(T!B) = 0. P(T!A) = 0.1x0.38 P’(A) = P(A!T) = 0.00928 .1 = 0.6 " 0.6.1 0.

P(D!WH) = 0.22x0.8 + 0.1x0.9x0.2 + 0.733 P( D) 0.2 (a) P(D!W) = P(D!WM)P(L)+P(D!WM)P(M)+ P(D!WH)P(H) = 0x0.06 . P (D!PM) = 0.05 = 0.02x0.5.2 where P denotes poorly constructed P(D!PL) = 0.2.22 Hence P(D) = 0.05 = 0. P(D!WL) = 0. 6% of the buildings will be damaged.05.20 P(H) = 1/20 = 0.2x0.2 " " 0.5x0.2 + 0.22 # 0.75 + 0.2 = 0.75 P(M) = 4/20 = 0.05x0. (c) P ( P!D) " P( D!P) P( P) 0.06 In other words.58 Given: P(L) = 15/(15+4+1) = 15/20 = 0.75 + 0. P(D!PH) = 0.02 (b) P(D) = P(D!W)P(W)+ P(D!P)P(P) But P(D!P) = P(D!PL)P(L)+P(D!PM)P(M)+ P(D!PH)P(H) = 0.05 P(P) = 0.1.9 P (D!WM) = 0.

2.25 + 0.145 (b) For a passenger boarding at Station 1.1)!0.3!0.59 (a) Note that any passenger must get off somewhere.6) miles = 11.5 + 0.4.3!0. P(D3) = P(D3 | O1)P(O1) + P(D3 | O2)P(O2) + P(D3 | O2)P(O2) = 0. Let Oi denote “Origin at station i” and Di denote “Departure at station i”.5 miles (c) Let E = “trip exceeds 10 miles”. These lengths have respective probabilities 0.1)!0. hence the expected trip length is (4!0.15 + (0.6!0.25 + 0.25 = 0.25 / 0.6!0.1 + 0. where i = 1.15 + 0.25 = 0.3.5 (d) P(O1 | D3) = P(D3 | O1) P(O1) / P(D3) = 0.3!0. 9 or 14 minutes for departures at stations 2. 3 or 4.517 . respectively.1.3 + 14!0.3 and 0.145 # 0.6.2. 0. P(E) = P(E | O1)P(O1) + P(E | O2)P(O2) + P(E | O3)P(O3) + P(E | O4)P(O4) = P(D4 | O1) P(O1) + P(D1 | O2) P(O2) + P(D1 " D2| O3) P(O3) + P(D2 " D3| O4) P(O4) = 0. his/her trip length would be 4.1 + 9!0.35 + (0. so the sum over any row is one.1!0.

(a) P(A!D) = = (b) P(D!C) = 0.48 0.02 # 0.1 P(D!A) = 0.36 = 0.02 # 0.36 P(A\$B!D) = 0.3 " 0.03x0. P (D!B) = 0.3.03.6 " 0.025 = 0.04 P( D!A) P( A) P( D!A) P ( A) " P( D!B) P( B ) " P( D!C ) P(C ) 0. P(C) = 0. P(B) = 0.48+0.02.6.84 .60 P(A) = 0.6 = 0.2.3 / 0.04 # 0.03 # 0.1 P(A\$B!D) = P(A!D)+P(B!D) Since P(B!D) = 0.

5x0.725 % 0.5 P(B) = 0.3 = 0.75 \$ 0.3 = 0.5 Where C denotes completion of project A on time P(C) = P(C!E)P(E)+P(C! E )P( E ) = 0.5 P(C!E) = 0.85 \$ 0.725 0.75x0.725 P( A( AB " AB)) P( A B " AB) # (d) # P( AB) P( AB) & P( AB) 0.654 0.6 P(C ) 0.5 =0.625 # P( B A) P( A) P( A " B) % P( AB) .5x0.25x0.5 + 0.2.5 P( AB( A " B )) P( AB) # # # # 0.625 (e) P ( E!C ) # P (C!E ) P ( E ) 0. P(C! E ) = 0.725 (b) P( AB A " B) # (c) P ( A AB " AB) # P( B A) P( A) 0.5 = 0.15 \$ 0.5 # # 0.3 P(A!B) = 0.5 =0.586 P( A " B) P( A " B) 0.25 P(B!A) = 0.15 (a) P(A"B) = P(A)+P(B)-P(A!B)P(B) = 0.61 P(winging project A) = P(A) = 0.3 – 0.75.5 + 0.85 \$ 0.5x0.

2 = 0.3 P(A!W) = 015 (a) P(W) = P(A)P(W!A) / P(A!W) = 0.8x0.0.06 = 0.7x0.3x0.2 = 0.46 Hence P(D) = 0.P(W!A)P(A) = 0.34 P(AW) = 0.2 = 0.4 = 0.06.15x0. P(D!AW) = 0.15 P(D! A W) = 0.14.14-0.3x0.4 . P( A W ) = 1-0.34-0.192 .4 (b) P(A"W) = P(A)+P(W).2x0.54 (c) P(A W " A W) = P(A"W)-P(AW) = 0.2 + 0.0.46 = 0.2 = 0.05 P(D) = P(D !A W ) P(A W )+ P(D! A W) P( A W)+ P(D!AW) P(AW)+ P(D! A W ) P( A W ) But P(A W ) = 0.2 P(W!A) = P(poor workmanship!poor aggregate) = 0.85x0.192 (e) P ( AW!D) # P( D!AW ) P ( AW ) 0.2x0.048 # # 0.34 + 0.25 P( D) 0.15 = 0.2.14 + 0.48 (d) Given P(defective concrete A W ) = P(D !A W ) = 0.2.8.3 / 0.3x0.05x0.06 + 0.54 .62 Given: P(A) = P(poor aggregate) = 0. P(D! A W ) = 0. P( A W) = 0.

2.1 / 0. N are mutually exclusive and collectively exhaustive.6!0.163 This is P*(S). (ii) (iii) Similar to (i). P(A | N) = 0.2. Since S.8. the updated probability of a small weak zone.6!0.05!0. L.8)!0.023 This is P*(L). P(N) = 0. The “updated” probability of A is calculated using P*(S) and P*(L).2!0.0233 + 0. Hence.1.3. P(L | E’) = P(E’ | L)P(L) / P(E’) = [1 – P(E | L)]P(L) / [1 – P(E)] = (1 – 0.2 / 0. P(A | L) = 0.135 (b) Let E denote “weak material Encountered by boring”.2 + 1!0. P*(A) = P(A | L)P*(L) + P(A | S)P*(S) + P(A | N)P*(N) Note that P(A | L).63 (a) Given probabilities: P(A | S) = 0. Hence P(A) = P(A | L)P(L) + P(A | S)P(S) + P(A | N)P(N) = 0.163) " 0.86 " 0.2.86 " 0.163 + 0.05!(1 – 0.3)!0.7 = 0.3)!0. P(E | N) = 0.05.1 + 0. P(A | S) all remains their original values as the settlement-soil condition dependence is not affected by the borings. hence P*(A) = 0. P(E’) = P(E’ | L)P(L) + P(E’ | S)P(S) + P(E’ | N)P(N) = (1 – 0 .0233 – 0. (i) The total probability of NOT encountering any weak material. we are given P(E | L) = 0.8)!0.6.2!0. the updated probability of a large weak zone in the light of new information.7 = 0. P(S | E’) = P(E’ | S)P(S) / P(E’) = (1 – 0.087 . P(L) = 0.86. P(A | N). applying Bayes’ theorem.7 and P(S) = 2P(L) lead to P(S) = 0.1 + (1 – 0.2 + 0. P(E | S) = 0.

3.01 + 0.004 < 0. P(DE!DL!DS) = 1 – P[ D E ! D L ! DS ] = 1 – P( D E D L DS ) by De Morgan’s rule.00502. the site in this case.01"0.1"0. P(D| A B) = 0. and D denote damage to dam. where i = E(arthquake). P(D) = P(D|A)P(A) + P(D| A )P( A ) = 0.02 – 0.2.01 + 0 = 0. as the new site has a higher risk.002)(1 – 0.02 = 0.5"0. (a) P(A!B) = P(A) + P(B) – P(AB) = P(A) + P(B) – P(A)P(B) since A and B are independent events = 0.4"0.001) # 0.1.99"0. B denote occurrence of earthquake in the respective cities. the new site is preferred since the yearly probability of incurring damages is about 20% lower.02 = 0.0298 (b) Given: P(D|A B ) = 0. S(oil being poor). P(D|AB) = 0. (d) Let Di denote “damage due to event i”.02 + 0.98 + 0.00502.00502 (c) Now that B has dropped out of the picture.01"0. = 1 – (1 – 0.01"0. L(andslide).5.3"0.64 Let A.006986 > 0.004) (1 – 0. due to . want: P(D) Theorem of total probability gives P(D) = P(D|A B )P(A B ) + P(D| A B)P( A B) + P(D|AB)P(AB) (note: P(D| A B ) = 0 so it is not included) = P(D|A B )P(A)P( B ) + P(D| A B)P( A )P(B) + P(D|AB)P(A)P(B) = 0. = 1 – P( D E )P( D L )P( DS ) one should move independence.

4 P( A) Hence.8 + 0.4x0. B.77 (b) P (C!A " B " C ) # (c) P(A B C " A B C " A B C) = P(A B C )+P( A B C )+P( A B C) But P ( B!A) # P( A!B) P( B) 0. C discover oil respectively P(A) = 0.288x0.52x0.0. P(B) = 0.4 = 0.0.2 = 0.6.2 P(A!B) = 1.6 + 0.72 0.65 A.8 + 0.28x0.2.397 . C denote company A.288 Hence P(A"B"C) = 1 – 0.48x0.26 P(A " B " C ) 0.8 = 0.48 \$ 0.6 = 0.6x0. B.6 # # 0. P(A B C " A B C " A B C) = 0.77 P(C (A " B " C )) P(C ) # # 0.48 C is statistically independent of B or A (a) P(A"B"C) = 1-P( A B C ) = 1-P( A B )P( C ) But P( A B ) = 1-P(A"B) = 1-P(A)-P(B)+P(A!B)P(B) = 1 .288x0.2x0. P(C) = 0.4.4 .

9.9"0. P(D | PB) = 0.9. P(D | A) = P(DA) / P(A) = [P(D | AB)P(AB) + P(D | AG)P(AG)] / P(A) (where DA = D(A#B) = DAB#DAG was substituted) (but P(D | AG) = 0) = P(D | AB)P(AB) / P(A) = P(D | AB)P(B | A) = 0. hence we may rewrite B as the union of two mutually exclusive events.3 + 0. Hence 20%"0.2"0.9"0.5 = 10% of morning flights will be delayed (note that good weather guarantees take-off without delay for morning flights). B = BA#BP. PB. there is 50% chance of having a delay.3 + 0.5"0.282 (b) Bad weather can happen during AM or PM hours.2.3.2 = 0.1"0.5"0.282 \$ 0. PG.1"0. When they do.8"0.7 + 0.7) / 0.8.1 (i.330 (c) Of all morning flights.5"0.7 = 0.5. (iii) P(A) = 0.7. in more formal notation. thus P(B | D) = P(BA | D) + P(BP | D) = P(BAD) / P(D) + P(BPD) / P(D) = [ P(D | AB)P(B | A)P(A) + P(D | PB)P(B | P)P(P) ] / P(D) = (0. AB.3 + 0.66 (a) Partition the sample space into AG.2 ! P(G | A) = 0.2"0.e. Theorem of total probability gives P(D) = P(D|AG)P(AG) + P(D | AB)P(AB) + P(D | PG)P(PG) + P(D | PB | PB) = P(D|AG)P(G|A)P(A) + P(D|AB)P(B|A)P(A) + P(D| PG)P(G|P)P(P) + P(D|PB)P(B|P)P(P) = 0"0. (ii) P(D | PG) = 0. P(D | AB) = 0. 20% will encounter bad weather.3"0. (iv) P(B | A) = 0. We are given (i) P(D | AG) = 0. Or. 10%) . P(B | P) = 0. P(P) = 0.9"0.3.1 ! P(G | P) = 0. (v) G and B are mutually exclusive and collectively exhaustive.

67 P(M) = 0.08 where W = carelessness of workers P(D!M W ) = 0.05x0.8 Assume M and W are statistically independent (a) P(D) = P(D !M W ) P(M W )+ P(D! M W) P( M W) +P(D!MW) P(MW)+ P(D! M W ) P( M W ) = 0.1x0.05x0.08 + 0. P(D! M W) = 0.05 where M = malfunction of machinery P(W) = 0.8x0.2. P(D!MW) = 0.8 .92 + 0.2x0.1.08 + 0 = 0.2.023 (b) P(W!D) = P(W M !D) P(WM!D) = (0.95x0.0032) / 0.0152+0.023 = 0.

8. Given: P (A | L) = 0.3) " 0. we have P(L | A ) " 0.314 " 0.1!0. even in L.98!0.2. P (B | L) = 0.955 (a) In terms of the ability to confirm actual existence of leakage. PL(A) = 0. B is better. with probability P( A B | L)P(L) = (1 – 0.e. B is better.7 + 1!0. hence the true probability P(A#B) = P(A#B | L)!P(L) = 0. there is a small piece of A’B’. P ( A | L ) = 1.8!0.8 and PL(B) = 0. . in terms of the ability to infer safety.314.44.44 " 0. In here. the only region in which A or B can happen. P (L) = 0.98.014 + 0.3 = 0. adding this to P( A B | L ) = 0. but this is only a relative probability with respect to L. and L denote that there was indeed leakage.9 – 0. Bayes’ theorem yields P(L | A ) = P( A | L)P(L) / P( A ) where P( A ) = P( A | L)P(L) + P( A | L )P( L ) = 0. In short. Note that.8)!(1 – 0.7 = 0.9 > P(A | L) = 0.e.7 = 0.1!0.2!0. P(B | L) = 0.189 Hence B gives a better assurance of non-leakage when it observes nothing. also these wells never false-alarm.7.8. while a similar calculation gives P(L | B ) = P( B | L)P(L) / P( B ) = 0. (a) Given A . B denote the events that these respective wells observed contaminants. hence P(L | A ) = 0. P( B | L ) = 1.68 Let A.318 as calculated in part (a).318 (b) (i) (ii) It will be convenient to first restrict our attention to within L.9. hence P( L | A B ) = P( A B | L )P( L ) / P( A B ) = 1!0.014.3 (since both wells never false alarm) gives the total probability P( A B ) = 0.3 / 0.9.686 A B is the event “both wells observed no contaminant”.8 + 0. Also. hence P (A#B | L) = PL(A) + PL(B) – PL(A)PL(B) = 0.7 / (0.3 = 0.7 + 1!0. i.9)!0.2!0. i.7 / 0.9 = 0.

05 = 0.2 since existence of pollutant B affects 20% of all people (even when A is absent) P(H | A B ) = 0 since there is no cause for any health problem.2 – 0. P( A B) = P(B) – P(AB) = 0.15 + 1!0. A B.2!0.05. The total probability of having a health problem is P(H) = P(H | AB)P(AB) + P(H | A B)P( A B) + P(H | A B )P(A B ) + P(H | A B )P( A B ) where P(H | AB) = P(H | A B ) = 1 since existence of pollutant A causes health problem for sure.5!0.05 + 0. P(H) = 1!0. also P(AB) = P(B | A)P(A) = 0.1 = 0. A B .5)!0.69 Let A and B denote having excessive amounts of the named pollutant. P(H | A B) = 0. Let us partition the sample space into AB.05 = 0.05.2.15 (this is clear from a Venn diagram). and H denote having a health problem due contaminated water. A B .1 = 0. P(A B ) = P( B | A)P(A) = (1 – 0.13 .

0041 + 0.2.4)2 = 0.70 (a) P(winning at most one job) = P(winning zero job) + P(winning exactly one job) = P( H1H 2 H 3 B1B2 ) + P( H1H 2 H 3 B1B2 ) + P( H1H 2 H 3 B1B2 ) + P( H1H 2 H 3 B1B2 ) + P( H1H 2 H 3 B1B2 ) + P( H1H 2 H 3 B1B2 ) = (0.4)6 + (0.4)4(0.081 = 0.6)4x3x(0.4)2 + (0.081 (b) P(win at least 2 jobs) = 1 – P(win at most one job) = 1 – 0.0461 .919 (c) P(win exactly 1 highway job)!P(win zero building jobs) = P[( H1H 2 H 3 )+P( H1H 2 H 3 )+P( H1H 2 H 3 )]!P( B1B2 ) = (0.6)x5 = 0.0768 = 0.

06 + 0.28 + 0.04 Check: 0.6\$0.2\$0.e.4 + 0.5 + 0.06 7 0.3.6\$0. Divide the sample space into A = 3.2\$0.n)P(A = n) n" 3.2 = 0.5 = n" 3.5 = 0.4 P(T = t) 0.3 = 0. 4 .22 P(T = 11) = P(B = 6)P(A = 5) = 0.2\$0.2 = 0.28 0.5 = P(B =4)P(A = 3) = 0.22 + 0.3 = 0.2\$0. The PMF is plotted as follows: 0.3 + 0.1 Total time T = A + B.5 + 0.4 P(T = 10) = P(B = 6)P(A = 4) + P(B = 5)P(A = 5) = 0. & c.04 = 1. which ranges from (3 + 4 = 7) to (5 + 6 = 11). events) ! P(T = 7) = # P(T = 7 | A = n)P(A = n) # P(B = 7 .2\$0. and A = 5 (m.2 = 0.28 P(T = 9) = P(B = 6)P(A = 3) + P(B = 5)P(A = 4) + P(B = 4)P(A = 5) = 0. A = 4.06 Similarly P(T = 8) = P(B = 5)P(A = 3) + P(B = 4)P(A = 4) = 0.2\$0.6\$0. 4 .e.22 0.04 8 9 10 11 t .

02) # 2] = 0. we see that P(X > 40) = Area of smaller triangle 2 % 70 \$ 40 ( =' * ! Area of larger triangle & 70 \$ 10 ) = 0.e.08 = 0.015 # 0.3.10)!(0. (a) P(lose money) = P(X < 0) = Area under the PDF where x is negative = 0.10)!(0. hence the conditional probability.52 [(70 .02 + (70 . money was made).2 Let X be the profit (in \$1000) from the construction job.1875 .015 Hence the required probability P(X > 40 | X > 0) is 0. respectively).02!10 = 0.02) # 2] = 0.015 # [10!0. P(X > 40 | X > 0) = P(X > 40 " X > 0) # P(X > 0) = P(X > 40) # P(X > 0) Let's first calculate P(X > 40): comparing similar triangles formed by the PDF and the x-axis (with vertical edges at x = 10 and at x = 40.2 (b) Given event is X > 0 (i.

3 " (a) Applying the normalization condition !f X ( x)dx = 1 #" 6 6 * x2 x3 ' x2 \$ c( x # )dx = 1 \$ c ( # % = 1 6 ) 2 18 & 0 0 18 = 1/6 \$ c.6) = [1 . whereas if C did not occur then overflow would correspond to X > 4.3.4)] = [FX(5) . 9 + 36 # 6 3 ! (b) To avoid repeating integration. which is 1 * x2 x3 ' FX(x) = ( # % 6 ) 2 18 & 2 3 = (9x – x )/108 (for x between 0 and 6 only) Since overflow already occurred.FX(5)]+0.FX(4)] / [1 .6 + [1 . the given event is X > 4 (cms).148 . hence the conditional probability P(X < 5 | X > 4) = P(X < 5 and X > 4) / P(X > 4) = P(4 < X < 5) / [1 – P(X .4 = (1 – 100/108)+0. let’s work with the CDF of X.6 + (1 – 80/108)+0.4 . 0.6.6 + P(X > 4)+(1 – 0. Hence the total probability of overflow is (with ’ denoting compliment) P(overflow) = P(overflow | C)P(C) + P(overflow | C’)P(C’) = P(X > 5)+0. where P(C) = 0.FX(4)] 2 3 2 3 2 3 = [(9+5 – 5 ) – (9+4 – 4 )]/ [108 – (9+4 – 4 )] = (100 – 80) / (108 – 80) = 20/28 = 5/7 . If C indeed occurs. overflow means X > 5. 0.714 (c) Let C denote “completion of pipe replacement by the next storm”.FX(4)]+0.

(2/3)4]2%0. E1 and E2 are m. obtain the PDF of X: fX(x) = dFX/dx = 4%104x-5 for x & 10.5 " FX(xm) = 0.2 severe snow storms in a year.e. and zero elsewhere.(2/3)4]%0.3 (inches) 3 (c) P(disastrous severe snowstorm) = P(X > 15) = 1 . let E0.5 " 1.1 \$ 0.3. Also.e.. is obtained by solving the equation that defines xm.(2/3)4 ]2 (statistical independence between storms). the total probability of N can be computed: P(N) = P(N | E0)P(E0) + P(N | E1)P(E1) + P(N | E2)P(E2) = 1%0. E2 denote the respective events of experiencing 0.(2/3)4. E1. and c.About 20% of snow storms are disastrous. Hence the expected amount of snowfall in a severe snow storm is ( E(X) = ) xf X ( x )dx '( ( = ) 4 % 10 4 x ' 4 dx 10 = ' 4 4 '3 ( 10 * x + 10 3 4 = % 10 4 \$ 13. definitely there won't be any disastrous one). xm.885 .(10/15)4] = (2/3)4 \$ 0.P(X ! 15) = 1 .1.4 + [1 . In each event. P(X ! xm) = 0. P(N | E2) = [P(X ! 15)]2 = [1 . P(N | E1) = P(X ! 15) = 1 .9 (inches) (b) First. (d) Let N denote "No disastrous snow storm in a given year".4 (a) The median of X.(10/15)4 = 1 .[1 .2 .(10 # xm)4 = 0. the (conditional) probability of N can be computed: P(N | E0) = 1 (if there's no severe snow storm to begin with. Noting that E0.5 " xm \$11.5 + [1 .

5 Let F be the final cost (a random variable).216395324 = 0. 1.25) = )f X ( x) dx 1.5 – 1) = 1.3.1 .5 1 while 2 E(X ) = 1.020382415 .216.5 1.5. 1. (a) To satisfy the normalization condition.25 1.0. 2 a x ' x \$1 hence a = 3/2 =1.25”.5 – 1. and C be the estimated cost (a constant).25 = -2 – (. i.2. hence X=F/C is a random variable.1.5 E(X) = 3 )x x 2 dx = +3 ln x .25 1. whose probability is * P(X > 1. a ) 1 a 3 3 ( " 3% dx ! & # ! 3 " ! 1 . we can determine the variance 2 2 Var(X) = E(X ) – [E(X)] 2 = 1.e.5 )x 1 2 3 dx = 3(1.143 .5 3 ( " 3% dx = & # 2 x ' x \$ 1.5 (b) The given event is “F exceeds C by more than 25%”.4) = 0.4 = ) (c) The mean.25C”.020382415 . “F > 1. x2 with these. “F / C > 1.e./X = 0. i.

3.106 0 3 (b) P(X < 3 | X > 2) = 1# 2/3 = 4/9 1 # 1/ 4 P( X ' 3 and X & 2) P(2 ' X ' 3) = = P ( X & 2) P ( X & 2) ! 2 2 dx x2 2 x 1# dx 8 0 ! # = 2 x 3 2 2 -x * 1# + ( .6 " 2 8 2 x x3 (a) E(X) = xf ( x)dx = x dx + x 2 dx = 8 24 x #" 0 2 ! ! ! 2 8 \$ 2 ln x 2 = 1/3 + 2(ln 8 – ln 2) % 3. 16 ) 0 2 = .

X > 30. i. and the probability of such an event is 4 4 (1 – p)4p = [1 . whose probability is P(X > 30) = 1 – P(X ! 30) = 1 – FX(30) 4 = 1 – [1 – (10/30) ] 4 = (1/3) = 1/81 " 0. let Y count the number of years in which failure occurs.9 lb/ft2.0123 # p Now for the first failure to occur in the 5th year.e.3. there must be four years of non-failure followed by one failure. hence the desired probability is P(Y < 2) = P(Y = 0) + P(Y = 1) n n–1 = (1 – p) + n(1 – p) p = (80/81)10 + 10\$(80/81)9\$(1/81) " 0.3 lb/ft2 and 11. respectively (as done in Problem 3-3-3). Y follows a binomial distribution with n = 10 and p = 1/81.994 .0117 (b) Among the next 10 years.(3/4)4 ] \$(1/3) " 0. (b) The event “roof failure in a given year” means that the annual maximum snow load exceeds the design value.7 (a) The mean and median of X are 13.

05 2 6 10 x (d) From the CDF. Hence E(X) = 1#0. 0. 8 with respective probabilities 0.8 (a) Note that we are given the CDF.V.4 + 4#0.4. FX(x) = P(X ! x) and it is a continuous function.4) / (6 – 2) " xm = 2 + 2 = 4 (c) The PDF is the derivative (i.3. it is seen that one may view X as a discrete R.4 (b) The median is defined by the particular x value where F(x) = 0.4) / (xm – 2) = (0.2 0. hence it is the following piecewise constant function: f (x) X 0. 4. which takes on the values 1.1 0.4 = 0.05 0.5 – 0.2.8 – 0.e. Hence P(2 ! X ! 8) = FX(8) – FX(2) = 0.5. 0. precisely. slope) of the CDF. the constant slope there gives (0.4 (recall probability = area of each strip) for calculating E(X).4 = 4.4 (integration would give the same result) .6 – 0.2 + 8#0. which occurs somewhere along x = 2 and x = 6.

0 for s & 0 \$ ! s2 s ( for 0 ' s & 12 f S ( s ) ) #! 288 24 0 for s % 12 " s is where fS has a maximum. the total probability of failure is P(S > R) = P(S > R | R = 10)P(R = 10) + P(S > R)P(R = 13) = [1 – FS(10)].0.074074074.3.3 3 2 = [1 – (-10 / 864 + 10 / 48)].s /144 + 1/24 = 0 * the mode ~ s = 6.7 / 0. - E(S) = + 12 + sf ( s ) ds ) ( s ) . The mean of S.9 (a) Differentiating the CDF gives the PDF.7 + [1 – FS(13)].- 0 = .0519 .0.3 = 0.7 + [1 – 1]. (12 4 ) 12 3 ( = = .0. 288 3 . s3 s2 ( )ds 288 24 .0.18 + 24 = 6 4 . 24 (b) Dividing the sample space into two regions R = 10 and R = 13.0. hence setting its derivative to The mode ~ s )=0 fS’( ~ ~ * .

0.1 which is a straight line segment decreasing from y = 0.5 ! xm = 20(1 – 0. comparing the two similar triangles which have area ratio 2:1.005x + 0.10 (a) The only region where fX(x) is non-zero is between x = 0 and x = 20.3. where fX(x) = F’X(x) = .1 (at x = 0) to y = 0 (at x = 20) fX(x ) 0 xm 2 x (b) The median divides the triangular area under fX into two equal parts.5 (20 – xm) / 20 = (1 / 2) 0.858 .5 ) " 5. one must have 0. hence.

only one of them exceeding 4mm has the binomial probability (where n = 4.* = 71 / 24 2 2. 8 + 2 0 2 (c) P(X < 4) = 1 – P(X > 4) Where P(X > 4) is easily read off from the PDF as the area (5 – 4)(1/4) = 1/4. / x3 .25 2 0.75 (b) E(X) = 1 1 (d) A vertical line drawn at the median xm would divide the unit area under fX into two equal halves.25) 3 440.* 0 . x( x / 8)dx 0 x(1 / 4)dx ) . hence P(X < 4) = 1 – 1/4 = 3/4 = 0.422 .1/ 0 x 5 2 0 2 2 2 x 5 5 5 / x2 . the right hand rectangle having area 0.96 (mm) . sketches follow: 2' x&5 x%5 0 \$ ! x 2 / 16 ! (a) Integration of the PDF gives FX ( x) ) # ! x / 4 ( 1/4 !" 1 fX(x ) 1/ 1 FX(x ) (2.11 x'0 0' x&2 . p = 0.5 = (5 – xm)(1/4) 3 xm = 5 – 4(0. 24 + 0 .3.5) = 3 (mm) (e) Since each of the four cracks has p = 0.75 40.25 probability of exceeding 4mm (as calculated in (c)).

3.12
x

(a) For 3 ! x ! 6, FX(x) =

24

\$t

3

dt # 4 / 3 " 12 / x 2 ; elsewhere FX is either 0 (for x < 3) or 1 (for x >

3

6)
FX is a parabola going from x = 3 to x = 6, followed by a horizontal line when plotted
6

(b) E(X) =

24

\$ x( x

3

)dx = 24(1/3 – 1/6) = 4 (tons)

3

2

2

(c) First calculated the variance, Var(X) = E(X ) – [E(X)] =

6

\$x

2

(

3

24
)dx - 16 = 24(ln 6 – ln 3) =
x3

24 ln 2 - 16
% c.o.v. =

'
24 ln 2 " 16
#
( 19.9%
&
4

(d) When the load X exceeds 5.5 tons, the roof will collapse, hence
P(roof collapse) = P(X > 5.5)
= 1 – P(X ! 5.5)
2
= 1 – FX(5.5) = 1 – (4/3 – 12/5.5 ) ( 0.063

3.13
(a)

Since the return period, !, is 200 years, this means the yearly probability of exceedance (i.e.
encountering waves exceeding the design value) is
p = 1 / ! = 1 / (200 years) = 0.005 (probability per year)

(b)

For each year, the probability of non-exceedance is 1 – 0.005 = 0.995, while the intended
lifetime is 30 years. Hence non-exceedance during the whole lifetime has the binomial
probability,
30

0.995 " 0.860

3.14
Design life = 50 years
Mean rate of high intensity earthquake = 1/100 = 0.01/yr
P(no damage within 50 years) = 0.99
Let P = probability of damage under a single earthquake
(a)

Using a Bernoulli Sequence Model for occurrence of high intensity earthquakes,
P(quake each year) = 1/100 = 0.01
P(damage each year) = 0.01p
P(no damage in 50 years) = (1 – 0.01p)50 ! 0.99
Hence, 1 – 0.01p = (0.99)1/50 = (0.99)0.02 = 0.9998
or p = 0.02

(b)

Assuming a Poisson process for the quake occurrence,
Mean rate of damaging earthquake = 0.01p = 0.01x0.02 = 0.0002/yr
P(damage in 20 years) = 1 – P(no damage in 20 years)
= 1 – P(no occurrence of damaging quakes in 20 years)
= 1 – e-0.0002x20
= 1 – e-0.004
= 0.004

3.15
Failure rate = ! = 1/5000 = 0.0002 per hour
(a)

P(no failure between inspection) = P(N=0 in 2500 hr)

(0.0002 " 2500) 0 e #0.0002"2500
=
= e-0.5 = 0.607
0!
Hence P(failure between inspection) = 1 – 0.607 = 0.393
(b)

P(at most two failed aircrafts among 10)
= P(M=0)-P(M=1)-P(M=2)
=

10!
10!
10!
(0.393)0(0.607)10 (0.393)(0.607)9 (0.393)2(0.607)8
0!" 10!
1!" 9!
2!" 8!

= 0.0068 – 0.044 – 0.1281
= 0.179
(c)

Let t be the revised inspection/maintenance interval
P(failure between inspection)
= 1-P(no failure between inspection)
= 1-e-0.0002t = 0.05
-0.0002t = ln 0.95
and t = 256.5 \$ 257 hours

3.16
(a) Let C be the number of microscopic cracks along a 20-feet beam. C has a Poisson
distribution with mean rate !C = 1/40 (number per foot), and length of observation t =
20 feet, hence the parameter " = (1/40)(20) = 0.5, thus
P(C = 2) = e-0.5 (0.52 / 2!) # 0.076
(b) Let S denote the number of slag inclusions along a 20-feet beam. S has a Poisson
distribution with mean rate !S = 1/25 (number per foot), and length of observation t =
20 feet, hence the parameter " = (1/25)(20) = 0.8, thus
1 – P(S = 0) = 1 - e-0.8 # 0.551
(c) Let X be the total number of flaws along a 20-feet beam. Along 1000 feet (say) of such
a beam, one can expect (1/40)(1000) = 25 cracks and (1/25)(1000) = 40 slag inclusions.
Hence the mean rate of flaw would be ! = (25 + 40)/1000 or simply (1/40) + (1/25) =
0.065 flaws per foot, which is multiplied to the length of observation, t = 20 feet to get
the parameter " = 1.3. Hence
-1.3

P(X > 2) = 1 – P(X \$ 2) = 1 – e
# 0.143

2

(1 + 1.3 + 1.3 / 2!) = 1 - 0.857112489

(d) Thinking of beam rejection as “success”, the total number (N) of beams rejected
among 4 would follow a binomial distribution with n = 4 and p = answer in part (c).
Thus
+ 4(
P(N = 1) = )) && 4%0.142887511%0.8573 # 0.360
*1'

3.17
(a) Let N be the number of poor air quality periods during the next 4.5 months; N follows a
Poisson process with mean value (1/month)(4.5 months) = 4.5, hence
-4.5
2
P(N ! 2) = e (1 + 4.5 + 4.5 /2!) " 0.174
(b) Since only 10% of poor quality periods have hazardous levels, the “hazardous” periods (H)
must occur at a mean rate of 1 per month#10% = 0.1 per month, hence, over 3 months, H has
the mean
\$H = (0.1)(3) = 0.3
-0.3
" 0.259
% P(ever hazardous) = 1 – P(H = 0) = 1 - e
Alternative approach: use total probability theorem: although there is (1 – 0.1) = 0.9
probability of non-hazardous pollution level during a poor air quality period, during a 3month period there could be any number (n) such periods and the probability of nonn
hazardous level reduces to 0.9 for a given n. Hence the total probability of non-hazardous
level during the whole time is
'
'
'
e ) (1#3) (1 # 3) n
(3 # 0.9) n
-3 3#0.9
-0.3
=e e
0.9 n P ( N ( n) =
= e , hence
0.9 n
= e )3
n!
n!
n (0
n (0
n (0
-0.3
P(ever hazardous) = 1 – P(never hazardous) = 1 – e = 1 - 0.741 " 0.259

&

&

&

3.2 + (e #0.2) % 0. Hence P(2 out of 5 years are good) . hence.2.4 (disasters per year). respectively.536 (from (a)). we can let D be the combined number of earthquakes or/and tornadoes. P(F) = 1/5 = 0.3 (1 – 0. T.287 . with mean rate "D = "E +"T = 0. 2) (c) Let’s work with D as defined in (a). P(only one incidence of natural hazard) = P(D = 0)P(F) + P(D = 1)P(F’) –0.4 \$ P(good) = P(E = 0)P(T = 0)P(F’) = e-0.8 instead (b) In each year.8 % 0. due to statistical independence among E.4 =e #0.536 Note: alternatively. the (yearly) probability of flooding.1 + 0.4 P(D = 0)P(F’) = e #0.349 . P(good year) & p % 0.1. and compute -0.4)#(1 – 0.3 = 0.4 –0. They are both Poisson random variables with respective means 1 !E = "Et = #1 year = 0. F -0.3 10 years Also.18 (a) Let E and T denote the number of earthquakes and tornadoes in one year.1 e-0.5* = ++ (( p 2 (1 ' p) 3 % 0. !T = 0.2) = e #0.

0167x3 = 0.189 (b) Mean rate of fatal accident = 1/3x0.0489 .0167 per year P(failure between inspection) = 1-P(no fatal accident within 3 years) = 1-e-0.3.19 Mean rate of accident = 1/3 per year (a) P(N=0 in 5 years) = e-1/3x5 = 0.05 = 0.

3. With n = 5. X has a 1 "20 miles = 0.4. hence Poisson distribution with !X = 50 miles –0.16 * 2' .4 P(X # 1) = 1 – P(X = 0) = 1 – e = 1 – 0 670320046 \$ 0. and p = daily accident-free probability = P(X = 0) \$ 0 670.20 (a) Let X be the number of accidents along the 20 miles on a given blizzard day. we obtain + 5( P(Y = 2) = )) && p 2 (1 % p) 3 \$ 0.33 (b) Let Y be the number of accident-free days among five blizzard days.

hence –0.5. so the mean rate of fatal accidents is \$F = \$x"0.1. X has a Poisson distribution with 3 "2 months = 0.2 = 0. whereas –(3/12)(4) 2 [(3/12)(4)] / 2! P(2 accidents in 4 months) = e –1 = e /2! # 0.1 P(fatalities in two months) = 1 – P(F = 0) = 1 – e # 0.05 per month)(2 months) = 0.5 P(X = 1) = e " 0. hence !X = 12 months –0.3.5 # 0.184 No. P(1 accidents in 2 months) and P(2 accidents in 4 months) are not the same.095 . F has a Poisson distribution with mean !F = (0. (b) 20% of all accidents are fatal.05 per month Hence the number of fatalities in two months.303.21 (a) Let X be the number of accidents in two months.

5'(0.0817 Total cost (in million dollars) 5 + 10 = 15 5 + 20 = 25 20 + 10 = 30 20 + 20 = 40 Since the goal is to reduce the yearly flooding probability to at most 0.0. all these options will work but the top one is least expensive.P(AB) = P(A) + P(B) . Hence the optimal course of action is to change the return periods of A and B to 10 and 20 years.43 (c) Let !A and !B be the improved return periods for levees A and B. respectively.P(X < 2) = 1 . Using (i) from part (a).22 The return period ! (in years) is defined by ! = 1 where p is the probability of flooding per year.3.1 (probability per year) (a) P(town encounters any flooding in a given year) = P(A"B) = P(A) + P(B) . we construct the following table: New !A (and cost) New !B (and cost) 10 (\$5M) 10 (\$5M) 20 (\$20M) 20 (\$20M) 20 (\$10M) 30 (\$20M) 20 (\$10M) 30 (\$20M) Yearly flooding probability 1 1 1 = # \$ ! A ! B ! A! B 0.2 (probability per year) P(B) = 1/(10 years) = 0.28)4 ( 0.(1 . respectively.28)5 .28.0975 0. .f(0) .0.P(A)P(B) (! A and B s. p Therefore. Hence P(X & 2) = 1 .2 + 0.) = 1 !A # 1 !B \$ 1 (i) ! A! B = 0.15.28 (b) Let X be the total number of flooded years among the next half decade.1 .28)'(1 .130 0.i.1 = 0.145 0.2%0.f(1) = 1 . X is a binomial random variable with parameters n = 5 and p = 0.0. the design periods of A and B being !A = 5 and !B = 10 years mean that the respective yearly flood probabilities are P(A) = 1/ (5 years) = 0.

i.1875 e-0.3 e-0.23 Given: ! = mean flaw rate = 1 (flaw) / 50m2.3 \$ 0. with which we can calculate the new probability P(replacement) = 1 .f(1) = 1 .1875.e-0. which gives the expected replacement cost 3.7 replacements on average.3.0155. which gives rise to an expected replacement cost of 100%0.1875 .037%100 = 3.037 (b) Since the probability of replacement is 0.f(0) . X is Poisson distributed. ! = 1 (flaw) / 80m2 ' # = 15/80 = 0.1875 \$ 0.e-0.0.3 (flaws) (a) P(replacement) = 1 . . we would expect 0. f(x) = e "# #x x! with # = !t = 15 / 50 = 0. t = area of a panel = 3m x 5m = 15m2 Let X be the number of flaws found on area t.7%\$5000 & \$18500 (c) For the higher-grade glass.3 .037 and there are 100 panels.0155%\$5100 \$ \$7920 (assuming each higher grade panel is also \$100 more expensive to replace) We can now compare the total costs: Let C = initial cost of old type panels Old type: Expected total cost = C + Expected replacement cost = C + \$18500 New type: Expected total cost = C + Extra initial cost + Expected replacement cost = C + \$100%100 + \$7920 = C + \$17920. which is less than that of the old type. ( the higher grade panels are recommended.P(0 or 1 flaw) = 1 .e.0.

0.1 . We can use the binomial distribution to compute P(at most 1 truck overloaded among 5) = b(0. number of trucks & n = 5. the average number of trucks is # = !t = 1(30 = 30 (trucks).9 x e ) 30 x! x *0 + -30 =e 27 x .95 .0.5e-5 % 0. During t = 30 minutes.1)4'(1 . hence with t = 5 minutes " # = !t = 5 (trucks) Hence P(X \$ 2) = 1 .24 (a) Let X be the number of trucks arriving in a 5-minute period.1) % 0.3.0.P(X < 2) = 1 .P(X = 0) .92 (c) We will first find probability of the compliment event. with the (Poisson) probability f(x) = #x e-# / x! but if none of them is to be overloaded (event NO). n.1)5 + 5'(1 . Any number (x) of trucks could pass.96 (b) Given: P(a truck overloads) & p = 0. x * 0 x! = e-30 e27 = e-3 . p) = (1 .P(X = 1) = 1 . Given: truck arrival has mean rate ! = 1 (truck) / minute. p) + b(1. 0.9x Hence the total probability of having no overloaded trucks during 30 minutes is P(NO | x = 0)P(x = 0) + P(NO | x = 1)P(x = 1) + …… + 30 x = .0.e-5 . hence the desired probability –3 P(any overloaded trucks passing) = 1 – e . the (binomial) probability is P(NO | x trucks pass) = 0. n.

hence for an observation period of t = 10 (years). Hence P(two tornado years in three years) & 3' +2.25 (a) Let X be the number of tornadoes next year. hence P(next year will be a tornado year) = P(X \$ 1) = 1 – P(X = 0) -# = 1 – e = 1 – 1/e % 0.3. expected) number of tornado years over a ten-year period is np % 6. with n = 3 and p = yearly probability of having any tornado(es) = 0.632 (b) This is a binomial problem.632)1 % 0. p = 0. = ) * 0.441 (c) (i) (ii) The tornadoes follow a Poisson model with mean rate of occurrence ! = 1 (per year). hence the mean (i. the expected number of tornadoes is # = !t = 10 Number of tornado years follows a binomial model with n = 10. != 20 years t = 1 year " # = !t = 1.6321.632 2 (1 ( 0.32 .632120559. X has a Poisson distribution with 20 occurences = 1 (per year).e.

3. X has a Poisson distribution with ! = (1/50 years)(20 years) = 0. Y has a binomial distribution with parameters n = 3.4)(e0. But since any number of strong earthquakes could happen during the next 20 years.4)(0. p = 0. hence P(X " 1) = P(X = 0) + P(X = 1) = e-0.938 (b) Let Y be the total number (maximum 3) of bridges that will collapse during a strong earthquake.769 .343)x [(e-0.343)(0.3)3 = 0. & P(no bridge collapses) = % P(no bridge collapses | x earthquakes occur in 20 years)P(X = x) x '0 & = % (0.441 (c) During a strong earthquake.3.4 (1 + 0.4) # 0.26 (a) Let X be the number of strong earthquakes occurring within the next 20 years.1372) = e-0.73 = 0.343.3\$(1 – 0.2628 # 0. we need to compute the total probability that there is no bridge collapse (with X as defined in part (a)).4x) / x!] x '0 = (e-0. hence the required probability is P(Y = 1) = 3\$0.4.4) & % x '0 [(0. the probability that all 3 bridges survive is (1 – 0.4)]x / x! = (e-0.3)2 = 0.

4! -0. i.4 – 0.594 (b) For each excavation that takes place. hence we have the total probability % \$ P(no damage | x excavations) P(x excavations) x &0 % = \$ e '! ! x x! (0.449 -0.16 = 0. the desired probability is P(D1# D2) = P(D1) + P(D2) – P(D1D2) = P(D1) + P(D2) – P(D1 | D2) P(D2) = P(D1) + P(D2) – P(D1 ) P(D2) = 0. but there must be no damage no matter what x value. the pipeline has 0.36 = 0.64 Alternative method: Let Di denote “damage to pipeline in i-th excavation”. Experimentally. Hence P(any damage to pipeline | X = 2) = 1 – P(no damage | X = 2) = 1 – 0.64 (c) Any number (x) of excavations could take place.6 x &0 % = e-! \$ x &0 -0. Since 40% of all excavations are damaging ones.4)(1/50) = (1/125) (damaging excavations per mile) .42 = 0.8 – 0. damaging excavations must also occur as a Poisson process. hence P(at least two excavations) = 1 – P(X = 0) – P(X = 1) = 1 – e-! (1 + !) = 1 – e-2(3) " 0.8 =e Alternative method: recall the meaning of ( in a Poisson process—it is the mean rate. X has a Poisson distribution with mean ! = (1/50 miles)(100 miles) = 2.4 + 0.6! ) x -! 0.6 probability of having no damage.4(2) =e =e " 0. and hence (1 – 0. it would be determined by (& nE N where nE is the number of excavations observed over a very large number (N) of miles of pipeline. the true proportion of occurrence over a large period of observation.6! =e e x! x 0.3.4n E (D & & 0. hence N (D = (0.62 = 1 – 0.27 (a) Let X be the total number of excavations along the pipeline over the next year.4) = 0.e. but with the “diluted” mean rate of 0.4( .4 probability of getting damaged.

)(100 mi.8 =e ! 0.Hence P(no damaging excavation over a 100 mile pipeline) –(1/125 mi.449 .) = e–100/125 = e–0.

3. hence 12 inches Hence P(1 flaw in 3 connections) = P(1 flaw in 90 inches of weld) = P(F = 1) –0.25 P(acceptable connection) = P(X = 0) = e # 0. hence the desired probability is P(Y \$ 2) = P(Y = 2) + P(Y = 3) = 3 p 2 (1 % p ) + p 3 # 0. 4. as an unacceptable connection does not necessarily contain only 1 flaw—it could have 2.75 # 0. hence 12 inches –0.75.875 (c) The number of flaws (F) among 90 inches of weld has a Poisson distribution with 0 .1 !X = "30 inches = 0.75 =e "0.1 !F = "90 inches = 0.354 Note: “1 flaw in 3 connections” is not the same as “1 unacceptable connections among 3”.778800783. 3. .779 (b) Let Y be the number of acceptable connections among three. etc. X has a Poisson distribution with 0 . Y has a binomial distribution with n = 3 and p = P(an acceptable connection) = 0.25.28 (a) Let X be the number of flaws along the 30-inch weld connection.

165 = 0.29 Mean rate of flood occurrence = 6/10 = 0.6 " 3) 2 (0.8+1.62+0.02 = 0.6"3 \$ \$ ]e =[ 1! 2! 3! = (1.98 .6x0.02 P(treatment plant will not be inundated within 5 years) = P(no occurrence of inundating flood) = e-0.3.6 per year (a) P(1 ! N ! 3 within 3 years) (0.972)x0.6 " 3) 1 (0.6 " 3) 3 #0.725 (b) Mean rate of flood that would cause inundation = 0.

Hence. thus P(A % 3) = 1 – P(A = 0) – P(A = 1) – P(A = 2) –3 2 = 1 – e (1 + 3 + 3 /2) \$ 0.it compares how frequently I occurs relative to N.01 : 0. in which “success” means “injury” and “failure” means “non-injury”. and 2 2 P(2 successes among 2 trials) = p = (1/6) \$ 0.05/mi. thus P(N) = 5/6 and P(I) = 1/6 whenever an accident occurs. accident of either type.05 -. hence A has a Poison distribution with "A = (0.257 (b) A.5 2.028 .52 / 2 \$ 0.01/mi)(50 mi) = 0.577 (c) Let us model this as an n = 2 binomial problem. How do we determine p? Consider the physical meaning of the ratio !I : !N = 0. with !I = 0.3.06.30 Given: both I and N are Poisson processes. along a 50-mile section of the highway. "N= (0.5 (a) P(N = 2) = e # " N ("N )2 / 2! = e–2.01/mi and !N = 0. has the combined mean rate of occurrence !A = !I + !N = 0. where each “trial” corresponds to an accident.05/mi)(50 mi) = 2.06/mi)(50 mi) = 3. Therefore. p = 5/6. I and N have Poisson distributions with respective means "I = (0. hence their relative likelihood must be 1 to 5.5.

the (Poisson) probability of having no thunderstorm in December is -173 / 126 = 0. Having such a “success” in 2 out of the next 5 years has the (binomial) probability * 5' 2 (( %% p (1 – p)3 = 5!/2!/3!# 0.0. at a combined mean rate of !X = (173 + 840)/(21#6) = 8. thus \$X = (!X )(1 month) = 8.040 )/4! " 0.31 (a) Let W and S denote the number of winter and summer thunderstorms.3.253)3 " 0.2532#(1 .03968254 (storms per month).253. with winter and summer thunderstorms (X) that could happen simultaneously.040 4 P(X = 4|t = 1 month) = e (8. Their mean occurrence rates (in numbers per month) are estimated based on the observed data as (i) !W " (173)/(21#6) " 1. P(W = 0 | t = 1 month) = e let’s call this p. hence -8.37.267 ) 2& . respectively.056 (c) In any particular year.03968254. (ii) !S " (840)/(21#6) = 6.67 (b) Let us “superimpose” the two months as one.

\$ P(more than two defects) = 1 – P(two or less defects) = 1 – [P(D = 0) – P(D = 1) – P(D = 2)] –3 2 = 1 – [e (1 + 3 + 3 /2!)] = 1 – 0. Since only 20% of defects remain after inspection.95)/5 = 0.2 = "0. thus if we require P(0 defects) = 0. only about 1% of defects can go on undetected. for 1000 meters of seams.95) \$ p = – ln(0. i. we have 1 !R = !D"0.2 = 0.001/m)(3000m) = 3. then the mean rate of undetected defects is !U = !D"p.32 (a) (b) (c) Let D denote defects and R denote defects that remain after inspection.577 Suppose the allowable fraction of undetected defects is p (0 < p < 1).95 –5p \$ e = 0. undetected defects have a mean of # = !R t = (0.001 per meter 200 meters For t = 3000 meters of seams.0103. hence.3.95 \$ –5p = ln(0. the (Poisson) mean number of defects is !U"1000m = (1/200m)(p)(1000m) = 5p. .423 % 0.e.

5 =e ) (a) Let M denote the event that John misses his connection. let X be the delay time of the flight in hours. the flight will have a delay time of at least 0.7 = e– 1/0. or the opposite.5) –1/0. Hence the desired probability is P(X > 1 | X > 0.e.5) = P(X > 1 and X > 0. and the probability of such an event is – 2/0.135"0. for them to both miss their connections the flight must experience a delay of more than two hours.3 + P(X > 2)"0. Also.3 and P(J2) = 0. Note that -x/0.053 (b) Regardless of whether John has a connection time of 1 hour and Mike has 2.5 P(X > 2) = e # 0.3 + e– 2/0. the flight delay time exceeded his scheduled time for connection.7.5 " 0.368 . P(M) = P(M | J1)P(J1) + P(M | J2)P(J2) = P(X > 1)"0.5 –2 –1 =e /e = e / e = 1/e # 0.5) = P(X > 1) / P(X > 0.5) / P(X > 0.3 + 0. respectively.5 P(X > x) = 1 – P(X ! x) = 1 – F(x) = 1 – (1 – e -x/0.5/0.3.5 –0.5 hours when it arrives.0183"0.7 = 0.33 Let J1 and J2 denote the events that John’s scheduled connection time is 1 and 2 hours. Using the theorem of total probability.5 " 0. where P(J1) = 0.7 # 0.018 (c) Since Mary has already waited for 30 minutes. i.

665 &1000 # \$\$ !! 20 % " (b) P(at least two bars will fail) = 1 – P(at most 1 bar will fail) = 1 – P(x=0) – P(x=1) & 20 #& 980 # \$\$ !!\$\$ !! 1 19 % "% " = 1 – 0.058 (c) Let n be the required number of bars tested P(all n bars will pass test) & n #& 980 # \$\$ !!\$\$ !! % 0 "% n " ! 0.277 = 0. hence. m = 20.10 = &1000 # \$\$ !! % n " n ' 107.665 – 0.34 1000 rebars delivered with 2% below specification (a) 20 rebars are tested. N = 1000.712 ( 108 bars required .665 – 1000 & # \$\$ !! % 20 " = 1 – 0.3. n = 20 P(all 20 bars will pass test) = P(zero bars will fail test) & 20 #& 980 # \$\$ !!\$\$ !! % 0 "% 20 " = = 0.

P(none of the four gaps were large enough) –2. being able to cross).5 a long enough gap that allows crossing. which is exponentially distributed with a mean of E(T) = (1/10) minute = 0.1).e.290 . there is (1 – e–2. G follows a geometric distribution. if N is the number of gaps one must wait until the first “success” (i.35 (a) “A gap larger than 15 seconds” means that there is sufficient time (15 seconds or more) between the arrival of two successive cars.1 minute = 6 seconds.5 3 –2.18 gaps before being able to cross.e. –15/6 –2. Hence the desired probability is –2. which has probability e . n–1 –2.5 2. the probability is (1 – e ) .063 More formally.5) chance that it is not long enough for the driver to cross. followed by –2.3.5 3 For such an event to happen 3 times consecutively.082 (b) For any one gap. i.5 P(N = n) = (1 – p) p where p = e (c) Since the mean of a geometric distribution is 1/p (see Ang & Tang Table 5.5 (1 – e ) (e ) " 0.5 4 = (1 – e ) Hence –2. (d) First let us find the probability of the compliment event. he has to wait a mean number of –2. Let T be the (random) time (in seconds) between successive cars. –2.5 4 P(cross within the first 4 gaps) = 1 – (1 – e ) = 1 – 0.5 hence P(T > 15) = 1 – P(T ! 15) = e =e " 0.70992075 " 0.5 1/p = 1/ e =e " 12.

25 ! 2.71)/2.4)] / [1 . the conditional probability P(X > 77.#((77 .#((72 ..4)] =0.8446 .0. X > 72) = P(X > 77 and X > 72) / P(X > 72) = P(X > 77) / P(X > 72) = [1 .5)] = [ 1 .#(0.5 (a) Let X be the length (in micrometers) of any given crack.5 ) (b) Given that 72 < X .71)/2.#(2.36 Standard deviation of crack length= 6.2 = 1.3.024 . \$ 74 " 71 % = 1 " # 1.2 P(X > 74) = 1 " # & * + ' ( 2.5)] / [1 .0082 / 0.

707 + P(C<90-15)x0.293 90 " 80 ) & 0.707 (b) P(Project completed on target) = P(T) = P(T%E)P(E) + P(T% E )P( E ) = P(C<90)x0.293 = 0.8412 = 0.586 .707 + 0.655x0.37 (a) P(activity C will start on schedule) = P(A<60) ! P(B<60) 60 " 50 60 " 45 = #( )# ( ) \$ # (1)# (1) 10 15 = 0.293 = #( 25 = 0.421x0.3.2) & 0.707 ' # ( "0.

400)/120) = 1 .00621 or.5) or. % T = 0. T1 = Traffic volume after 10 years.!((350 .3.400)/120) = 1 – 0.!((C . C = 300 + 400 = 700 .99379 = 1.60) (a) P(T>350) = 1 .200)/60) = 1 . T1 is N(400.!((350 .662 (c) C = Capacity of airport after 10 years.5) = 0. P(T1>C) = 0. 120) P(T1>350) = 1 . " T1 = 200 + 0. T = The present traffic volume = N(200.10#200#10 = 400 \$ T = \$ T = 60/200 = 0.!(0.5 or.3#400 = 120 1 So.38 Let.42) = 0. (C . 1 .400)/120 = 2.!(2.3 1 Hence.!(2.00621 (b) Let.

5) \$ 0.8 + 0.39 A = volume of air traffic C = event of overcrowded (a) P(C!N-S) = P(A>120) 120 " 100 ) \$ 1 " # (2) \$ 0.3.2 = 0.02275 = 1 " #( 10 (b) P(C!E-W) = P(A>115) 115 " 100 ) \$ 1 " # (1.02275x0.0668 = 1 " #( 10 (c) P(C) = P(C!N-S)P(N-S)+P(C!E-W)P(E-W) = 0.0315 .0668x0.

2 ! x !X 2 % !x ) \$ 0.5 . = Hence. X ~ N( ! X .5) \$ 1 % & ( 2. P( X # 2) \$ & ( or and "X \$ 0. " X \$ 0. " X ).2 ( 1. Given probability: P(X # 2) = 0.2 ! x 2 % !x \$ & %1 (0.95 Also.5 P( X ' 2.5 % 1.95 0.2 ! x ! x \$ 1.40 Let X be the settlement of the proposed structure.00047 0.95) \$ 0.v. since we’re given the c.5 ) \$ 1 % & (3.95 \$ 1.645 0.063) \$ 0.3.o.2 .

X must be above 70 but below 100.8.41 (a) Let X be her cylinder’s strength in kips.309 " 0.5)] = 0. hence P(70 < X < 100) 100 ! 80 70 ! 80 " #( ) ! #( ) 20 20 " # (1) ! # ( !0. D is normally distributed with a mean of 'D = 'Y – 'X = 80.8 > 0 This suffices to conclude that the guy’s cylinder is more likely to score higher.159 / 0. Therefore.5 &D hence it is more likely for the guy’s cylinder strength to be higher than the girl’s. To be the second place winner.514 (c) Let Y be the boyfriend’s cylinder strength in kips.532 (b) P(X > 100 | X > 90) = P(X > 100 and X > 90) / P(X > 90) = P(X > 100) / P(X > 90) = {1 – #[(100 – 80)/20]} / {1 – #[(90 – 80)/20]} = [1 – #(1)] / [1 – #(0.5) " 0.01%80 = 80.8. Y ~ N(80.309 \$ 0.8 P(D > 0) = # ( ) ( 0. &Y) Let D = Y – X.8 – 80 = 0.3. .841 ! 0. 0 ! 0. Mathematically. which has a mean of 1.

727 .266x0.374 + 4 = 5.2) 6!4 (a) P(H>6) = 1 ! " ( (b) Design requirement is P(no exceedance over 3 years) = 0.8 = (1-p)3 Where p = p(no exceedance in a given year) 3.2 ) # 1 ! " (0.266 Hence.2 h = 3. p = 1-(0.2x0.2 " -1(0.8)1/3 = 0.1064x3 =0.1064 Hence mean rate of damaging wave height = 0.3.625) # 0.2 m (c) Probability of damaging wave height in a year = 0.1064 per year P(no damage in 3 years) = P(no damaging wave in 3 years) = e-0.0717 3.4 = 0.0717) + 4 = 3.0717 That is.3.42 H = annual maximum wave height = N(4. " ( h!4 where h is the designed wave height ) # 0.

P2(violation) = 1 – 1 =0 < P1(violation).0137.075) = % ') * = %( 3.03.43 Let X be the daily SO2 concentration in ppm. . the probability of X being under 0. Hence 0. hence + 0.9863 = 0.012/ 7 ).04) = % ') * & % " 2.4!0. hence + 0.986 .205# & 0. i.. On any one day.012 / 7 . /criteria 1 is more likely to be violated. where X ~ N(0.03. 0.3. i.075 \$ 0.012.04 \$ 0. X is normal with mean 0.012 . more strict.75) -.e. Thus the weekly mean. P1(violation) = 1 – 0.03 ( P(X < 0. P(no violation) = P(0 or 1 day being under 0.03 ( 1.03). whereas 2.075 is 0. X ~ N(0.e. P( X < 0. 0.03 and standard deviation 0.075 out of 7 days) 7 6 = 1 + 7!1 !(1 – 1) =1 .

\$ 1 # % ! 2" \$ 1 # 0.933 (c) Now.995 which is larger than the answer in (b).99 / p = 0. while p = 0. this is a better option.44 Let F be the daily flow rate. (d) In this case we work backwards—fix the probability of violation. X follows a binomial distribution with n = 3 and p = 0. we're given F ~ N(10. P(not charged) = P(X = 0 or X = 1) = P(X = 0) + P(X = 1) = (1 .2) (a) P(excessive flow rate) = P(F > 14) ' 14 # 10 ( \$ 1# %) * + 2 .p)3 = 0. but recall from part (a) that p is obtained by computing P(F > 14) / 0. Since the non-violation probability is larger.712 / 0 F .0.p)4 .02275 (b) Let X be the total number of days with excessive flow rate during a three-day period.003344 =1 # % )' 14 # 0 F *( + 1F .p.01 / P(non-violation) = 0. 14 # 0 F =% -1 ! 0.02275 (probability of excessive flow on any given day).02275 remains the same.(1 .p)5 + 5.0.3. with n changed to 5. and determine the required parameter values.p)3 . hence P(no violation) = P(zero violations for 3 days) = P(X = 0) = (1 . hence.00334.99 / (1 .99666 " \$ 2.97725 & 0.58 2 . We want P(violation) = 0.8.

0.034 + . ( % " 0. as follows: ' ln 20 .45 (a) The "parameters" ! and " of a log-normal R.4.3.779895 % 0.! * 2) -2) . hence we seek the conditional probability P(100 < T < 101 | T > 100) = P(100 < T < 101 and T > 100) / P(T > 100) = P(100 < T < 101) / P(T > 100) = P(ln 100 < ln T < ln 101) / P(ln T > ln 100) ' ln101 .779895 1 .406) % 0. Probability calculations concerning T can be done through X.385 and ! 0 4. ) ( #+ 2 ! % ln # Substituting the given values .T / "2 2 \$T % 0.0.0.V.308 The importance of these parameters is that they are the standard deviation and mean of the related variable X / ln(T). we find #T " 2 = 0.787468 . while "a severe quake occurs over the next year" is the event "100 < T < 101".! * ' ln100 . " " + ( + % ( ln100 ! ' * 1. are related to its mean and standard deviation # and \$ as follows: 2 '\$ * " % ln(1 & ) . #T % 80 . hence " 0 0.4.3078 * % 2 ( -3.385 ( + P (T 1 20) % 2 ) (b) (c) "T > 100" is the given event.14842 and ! = 4.307817. while X is normal.2) .00033 . .

074 Hence P(X > 100) = 1 – P(X % 100) = 1 – &( ln100 ' 4. hence the required probability is P(Y = 0) = 0. p = 0.972 .2 = 1 – &(2. The parameters of X are: " # ! = 0.9964).996024 # 0.004 (b) Let Y be the total number of days on which critical level is reached during a given week.46 (a) Let X be the daily average pollutant concentration. Y follows a binomial distribution with parameters n = 7.2. \$ # ln 60 – 0.654) = 1 – 0.074 ) 0.22 / 2 # 4.9967 # 0.004 (and 1 – p = 0.3.

222 1 % (( ) 0.953 ) 1 % ( (1.2936 H 1 2 ' ! ln & % \$ 2 ! 2.267 0.21 = ln 24 % 2.767) 0.47 \$ ! ln(1 " # 2 ) ! 0.2936 P( H * 27 + H * 24) P( H * 27) ! P( H * 24) P( H * 24) ln 27 % 2.0468 0.953 % ( 1 (0.2936 1 % (( .622) ! 0.677) 0.953 ) ! 1 % ( (0.3.953 (a) P(25 m long pile will not anchor satisfactorily) = P(H>25-2) = P(H>23) = 1 % (( (b) P(H>2T)H>24) = ln 23 % 2.2936 ! ! ! 0.

460172104 / [1 – )(– 0.716 .22 / 2 + ln100 ' 4.585 .355 + 0 ' 0. L is log-normal with parameters #L = 15/50 = 0.585 )] = [answer to (a)] / [1 – )( 0. with 2 2 1/2 2D = \$X – \$L = 0.2942 / 2 = 3.665 " 0.1 * 1 ' 0. X is log-normal with parameters !X " #X = 0.2. so it is again normal.48 (a) Let X be the pile capacity in tons.355 0 (c) P(X > 100 | X > 75) = P(X > 100 and X > 75) / P(X > 75) = P(X > 100) / P(X > 75) ln 75 ' 4.54 * 0.294 1 \$L = ln 50 – 0. / 0.338) = 0.427)] = 0. * 1 ' ) 0.2 = 0.P(X ( 100) * 1 ' ) % & .0219 1 P(D < 0) = ) % & .9096 " 0. 0.427) = 0.2 = 0.716 and 3D = (!X + !L ) = 0.506 (d) P(X > 100 | X > 90) = P(X > 100 and X > 90) / P(X > 90) = P(X > 100) / P(X > 90) ln 90 ' 4.46 / )(1. \$X " ln 100 – 0.2 / 0 (b) Let L be the maximum load applied.3.46 / 0.585 = [answer to (a)] / [1 – )( )] 0.338)] = 0.46 / 0.869 It is convenient to formulate P(failure) as P(X < L) = P(X / L < 1) = P(ln(X/L) < ln 1) = P(ln X – ln L < 0) but D = ln X – ln L is the difference of two normals.3 2 1/2 1 !L = [ln(1 + 0.017 * 0.46 P(X > 100) = 1 .3 )] = 0.46 / [1 – )(–1. * ) '2.46 / )(0.692 .

3.33 0.99 0) . X ~ LN(\$. P(X 1 V) = 0.22 / 2 2 Since ! and \$ are the standard deviation and mean of the normal variate ln(X).2 ( + ln V . !) where ! % ln(1 & .4. ) ( #+ 2 \$ % ln # - !2 2 (a) Let X be the maximum wind velocity (in mph) at the given city. i. i.4.99 ' ln V .062 P(X > 120) = 1 .e. = 0.2 V % 141( mph ) . 2 ) / .0 ) .4. thus \$ % ln # - !2 / ln (90) – 0.938/ 0.e.48 % 0 -1 2 0.538) = 1 – 0.48 * % 0.2.01. ' ln120 .99 3 % 2.01. P(X > V) = 0. 0.2 ( + (b) To design for 100-year wind means the yearly probability of exceeding the design speed (V) is 1/100 = 0.48 * = 1 – 0(1.49 2 '" * ! % ln(1 & ) . 0.

143 .o.3. t " 58 days (b) P(T < 58+7#T > 58) = 1% (c) I (0.9656 0. k = 16.048 P(at least one out of three will breakdown) = 1 – P(none of the three will breakdown) = 1 – (0. 16) ! 0.7 P(T ' 58) I u (0.4 & 58. c.4 1 k Let t be the required maintenance schedule interval Hence.50 T = time until breakdown = gamma with mean of 40 days and standard deviation of 10 days (a) k/! = 40.533 0.16) I (26.95 By trial and error.v.9520 0.16) P(T ' 65) 0.8435 0.01417 \$ 1% u \$ 1% u \$ 1% \$ 0 .95)3 = 0. t 40 50 60 58 57 Probability 0. P(T<t) ! 0.95 or Iu(0.16) 0.944 Hence.4 t.16) I u (23.4 & 65. = 0.25 = Hence.2. ! = 0.

776 .905. 8.16) 0.00327x2000.745 & & & & 0.81.51 The capacity C is gamma distributed with a mean of 2.v. of 35% k/! = 2500.35 Hence. 8. 8.16) = 0.o.312 x 1/20 = 0. 8.00327 % 3000.00327 P(C >3000"C > 1500) P (C # 3000) 1 \$ I u (0.0156 P(no damage over 50 years) = e-0.458 (c) P(at least 4 buildings damaged) = 1 – P(none of the the 4 will be damaged) = 1 – (0. ! = 0.688)4 = 0.0156x50 = 0. (a) (b) 1 k = 0.16.889 P(C<2000) = Iu (0.3.16) 1 \$ I u (9.312 Mean rate of damaging quake = 0. k = 8.500 tone and a c.00327 % 1500.16) 0.838 = P (C # 1500) 1 \$ I u (0. 8.16) I u (4.

275 .4 + 1x0.1875x0. P(T>5) = P(T>5!T1 = 2)P(T1 = 2) + P(T>5!T1 = 3)P(T1 = 3) = P(T2>3)P(T1 = 2) + P(T2>2)P(T1 = 2) = 0x1/4 + 1/4x3/4 = 0.3 + 0.52 Ti = time for loading/unloading operation of ith ship T = waiting time of the merchant ship (a) Given that there are already 2 ships in the queue.1875 from above Given 3 ships in the queue.2 = 0. P(T>5) = P(T1>5) = 0 Given 2 ships in the queue. P(T>5) = 0x0. Hence it is certain that P(T>5) = 1 In summary. P(T>5) = 0.1875 (b) P(T>5) = P(T>5!N=0)P(N=0) + P(T>5!N=1)P(N=1) + P(T>5!N=2)P(N=2) + P(T>5!N=3)P(N=3) Given 0 ship in the queue.3. the waiting time will be at least 6 days.1 + 0x0. P(T>5) = 0 Given 1 ship in the queue.

\$ u (0.95.41.8 (0.41 750 = 600+ P(V>1000) = 1.8 P(V>1000) = 1.q # 0.097 P(J) = 0.02 + 0.971 .3.98 For the parameters of Beta distribution for V.o.v.95) But u = (1000-600)/(1100-600) = 0.2x750)2 = (1100-600)2 2 (q " r ) (q " r " 1) Hence. 0.95) = 0. q (1100-600) q"r qr (0.41.097x0.20 A = accidenct (a) P(Jamming occurs on the bridge) = P(J) = P(J!A)P(A)+P(J! A )P( A ) = 1x0.53 Traffic volume= V = Beta between 600 and 1100 vph with mean of 750 vph and c.\$ 0. r # 0.98 = 0. of 0.02 + P(V>1000)x0. 0.

00007 .8" 2 2 6 ' \$ = 28x(0.54 By using the binomial distribution.2 " !0.2)2(0.8)6 = 0.3.294 By using the hyper geometric distribution. we need number of passing students in a class of 30 = 24 number of failing students in a class of 30 = 6 Hence P(2 out of 8 students will fail) ( 24 %( 6 % && ##&& ## 28 ) 15 ' 6 \$' 2 \$ = * 5852925 ( 30 % && ## '8\$ =0. P(2 out of 8 students will fail) (8% = && ##!0.

736 (b) For the total of 50 road graders.v.457x40. Number of graders with T<40 = 0. ! = 0. 16) = Iu(18.55 (a) T = trouble-free operational time follows a gamma distribution with mean of 35 days and c. 16) = 0.3.457 P(T>40) = 1-Iu(0.0216 & 109 ) 0.o. of 0.25 Hence. number of graders. 1 k = 0.027 & 1010 " 50 # \$ 10 % ' ( .021 = ' (' ( ) 1. k = 16.3.1x50 = 5 P(2 among 10 graders selected will have T<40) " 5 # " 45 # \$2%\$ 8 % 10 & 0.25 k/! = 35.

C " 5.198 By trial and error.7.5 / 6.198 P(C>5.0004 P(building survives 100 years) = P(no damaging earthquake in 100 years) = e-0.737 4 1 5 0 mean rate earthquake that causes damage to at most 3 structures = 0.5.0.4096+0.863 .5 0. 198 = \$ \$ 0.3.84) \$ 0.5 For lognormal distribution.0.23 and ( )' Since ! =ln 6.5 1 * +( ) 0.5 / # ) 2 # ) 6.7 \$ 0. # " x m & / # % & / x m \$ 1.96 (d) P(survival of at least 4 structures during the earthquake) 552 5 52 = 33 00.0.2 / 4 51 4 41 = 0.5) \$ P(C " 4) P (C " 4) ln 5.0004x100 = e-0.907 0 .2 0.198 P(C " 4 . and ( ) 1.00147x100 = e-0.5 / 6.5 = ln # * (1.0.3277 = 0.2 / 6 33 00.2x1/500 = 0. hence (a) (b) (c) P(Damage) = P(C<5.5 ! = ln 6.737x1/500 = 0.147 = 0.9929 1 * +( ) 0.C>4) = Mean rate of damaging earthquake = 0.5 * ln 6.5 * ln 6.8/ .5) P(C " 5.5 \$ 0.5 and standard deviation of 1.00147 P(at least four building will survive 100 years) = e-0.915 ln 4 * ln 6.04 = 0.5 ) \$ + (*0.8/ .5) = + ( ln 5.56 Seismic capacity C is lognormal with median of 6.

9556) " 0.22(1 – 0. and hence P( N B | N A ) = 220.955319) " 0.57 (a) Let A and B denote the pressure at nodes A and B.o. P( N B | N A ) = 220. and #A = ln(\$A) – ! A2 2 " ln(10) – 0. respectively.10.788 ( + (2.2 (small).2913 0. P( N B | N A )P( N A ) becomes 0.0021 (II) to change P( N B ) to 0.1 = 0. .) + 1. we get the new values of: #A !A 2.283 .0044 % Option I is better since it offers a lower probability of unsatisfactory water services than II.102(1 – 0.98934 hence the probability of unsatisfactory water services.2 0 1 0 1 (b) Let Ni denote the event of pressure at node i being within normal range.3. & ' & ' / .33 upper limit for Z 2.318 P(normal pressure at A) 0.1. we have ! A " 0.9 % P( N B ) = 0.4560 ) 0. / 0.22(1 – answer to (a)) = 0.955 ) P (6 * A * 14) ) + .9893) " 0.22(1 – 0.02 = 2.2 0. Given: P(NB) = 0.2 Hence P(unsatisfactory water services to the city) = P( N A 3 N B ) = P( N B | N A )P( N A ) = 0.05.B.282585093 % P(satisfactory performance at node A) .05 = 0. Since A is log-normal with mean = 10 and c. ln 6 ( 2.v.15: repeating similar calculations as done in (a). thus P( N B | N A )P( N A ) = 0.o.0089 (c) The options are: (I) to change the c.2.283 .v.( + . i = A. = 0. ln14 ( 2.15 lower limit for Z –3. of A to 0.

131 .0.Y ( x. while . y ) dxdy = 5 0 "" 0 1 = 1 1 " " (x 2 y # xy 3 )dxdy 0 0 1 2 3 3 ydy # y dy = 1/5 + 3/20 = 7/20 = 0.0.5 | X = 0. y ) = f X ( x) (6 / 5)( x # y 2 ) x # y2 =3 (2 / 5)(3 x # 1) 3x # 1 1 "f Hence P(Y > 0.01.5 # y 2 dy = (3/2.271 .5) '0.65 1 1 (c) E(XY) = 6 xyf X .5 " = 0.3.283.X = {E(X2) – [E(X)]2}1/2 = [(13/30) – (3/5)2]1/2 = 0.5 ( y | x * 0.01 0.35 50 50 " " + Cov(X. -XY = Cov( X . Hence the correlation coefficient.5 ( 0. 1 1 y3 & 6 6) !fX(x) = ( x # y 2 )dy = ' xy # \$ 5 5( 3 %0 0 2 (0 < x < 1) = (3x + 1) 5 " (b) fY|X(y|x) = f X .Y = .5) = Y |0.5 # 1 3 % 0.5)dy 0. X.5 y # \$ 1 .5 1 1 =3 ) y3 & 0 .Y ( x. Y ) .35 – (3/5)(3/5) = -0.271 / 0.58 (a) fX(x) is obtained by “integrating out” the independence on y.Y) = E(XY) – E(X)E(Y) = 0.Y = {E(Y2) – [E(Y)]2}1/2 = [(11/25) – (3/5)2]1/2 = 0.283 0 .

59 (a) Summing over the last row.30.10#2 = 0. Hence we have P(Y = 10 | X = 2) = 0.25#2 = 0. hence the marginal PMF of runoff Y is as follows: fY(y) 0. then (say) P(Y ! 20) should be the same as P(Y ! 20 | X = 2) = 0.15 + 0. hence X and Y are not s. so all those probabilities should now be divided by 0..15#2 = 0. reduced sample space corresponds to only the second column. not one.1 + 0.25 + 0.3.25 + 0.i.25 / 0.7. P(Y = 30) = 0. where probabilities sum to (0. 2nd & 3rd columns of the given joint PMF table. we use the probabilities in the X = 2 column. However.7. (d) Summing over each row.20. we obtain P(X ! 2 and Y > 20) = 0.20.2 (b) Given that X = 2.5 0.1 = 0. we obtain the (unconditional) probabilities P(Y = 10) = 0. the new.0 + 0. each multiplied by 2 so that their sum is unity.35 / 0.10 = 0.2 0.6 0. P(Y = 20 | X = 2) = 0. and hence the PMF plot: fY|X(y|2) 0. Hence P(Y ! 20 | X = 2) = 0.i.10 + 0.60.80 " 0.1 / 0.7 (c) If X and Y are s.5.10 + 0.2 10 20 30 y (e) Given that X = 2.25 + 0.5 + 0.3 0.20.5 = 0.2 10 20 30 y .5. P(Y = 30 | X = 2) = 0.10) = 0. P(Y = 20) = 0.5 = 0.50. P(Y ! 20) = 0.

y) all = 1!10!0.5!(2 – 2. P(X = 2) = 0.2!(30 – 20) = 40.6!20 + 0.15.15 + 1!20!0.5 Hence the correlation coefficient is E ( XY ) % E ( X ) E (Y ) \$= Var ( X ) Var (Y ) = 45. we calculate E(X) = 0.15!(1 – 2.5.2!30 = 20.2!(10 – 20) + 0.05 + 2!10!0.35!3 = 2.35!(3 – 2.46 ! 40 & 0. P(X = 3) = 0.15!1 + 0. we obtain the marginal PMF of X as P(X = 1) = 0.2) + 0.6!(20 – 20) + 0.5 % 2.2 ! 20 0.10 + 3!30!0.10 = 45. 2 2 2 Var(X) = 0.2) + 0.10 + 2!20!0. 2 2 2 Var(Y) = 0.25 + 3!20!0.46.5!2 + 0.2!10 + 0.35. similarly E(Y) = 0. Also.2) = 0.2. and results from part (d). E(XY) = " x#y#f(x. With these.35 .(f) By summing over each column.25 + 2!30!0.

").29). Y is LN with parameters ! being 2 and " being 0. x = g-1(y) = ln(y) dg )1 d 1 .6). the PDF fY(y) is = fX(g-1) = 1 dg )1 1 = fX(ln(y)) dy y 1 ( ln y ) ! % ) && ## 2' " \$ 2 e (non-negative y only) 2+ y" which. "). by comparison to Ang & Tang (3. and Y = eX. "). 0. .e. f X ( x) * 1 1 ( x )! % # ) && 2 ' " #\$ 2 e 2+ " To obtain PDF fY(y) for Y = eX. if X ~ N(!. = ln y * dy dy y dg )1 dy y = g(x) = ex. then Y ~ LN(!. for the particular case where X ~ N(2. let us first do it for the general case where X ~ N(!. Hence.4. i. i.4).e. one may apply Ang & Tang (4. expressed as a function of y. is exactly what is called a log-normal distribution with parameters ! and ".4. fY(y) = fX(g-1) where .1 Since it leads to an important result. 1 since y = eX > 0 y Hence.

let’s work with the CDF (which we can later differentiate to get the PDF) of Y: since Y cannot be negative. f Y ( y) 1 . hence when y < 0: FY(y) = 0 ! fY(y) = [FY(y)]’ = 0 But when y " 0.0 2y 2y exp(\$ ) 3 +m ma 2 y"0 y00 . a 3 . FY(y) = P(Y # y) = P( 1 mX 2 # y ) 2 2y 2y ) #X# m m = P( \$ * 2y ' * 2y ' % \$ FX ( \$ % ( m % ( % m ) & ) & = FX ( * 2y ' % ( m %\$0 ) & = FX ( * 2y ' % ( m % ) & = FX ( Hence the PDF. we know that P(Y < 0) = 0.2 To have a better physical feel in terms of probability (rather than probability density). fY(y) = d [FY(y)] dy * 2y ' d 2y % ( m % dy m ) & 8y 2y 1 exp(\$ ) = 2 3 ma ma + 2my = fX ( = 4 a3 2y 2y exp(\$ ) 3 +m ma 2 Hence the answer is /4 ..4.

973 P(at least one source not able to supply respective allocation) = 0.4.3x400)P(H>0.09962x0.667 + 0.5x0.5) = [1 ' ( ( = 0.5 3 58.15x400)P(F>0.977x1 = 0.667 \$ 0.667 P( S W ) P(W ) \$ 0.5 3 2 1 = [( (0)] & # [( (3.99968x0.5 & 0.55x400) P( S ) 220 ' 400 120 ' 200 60 ' 100 )] )][1 ' (( )][1 ' ( ( 40 40 15 = ( (2. ! T ) " T \$ 100 # 200 # 400 Where ! T \$ 15 2 # 40 2 # 40 2 \$ 58.027 .667 (c) P(W%S) = (d) P(all individual power source can meet respective demand) = P(N>0.5 (b) P(Normal weather) = P(W) = 2/3 P(Extreme weather) = P(E) = 1/3 P(Power shortage) = P(S) = P(S%W)P(W)+P(S%E)P(E) = P(T<400)x2/3 + P(T<600)x1/3 400 ' 400 2 600 ' 400 1 )] & # [( ( )] & 58.333 = 0.5 0.3 A = volume of air traffic C = event of overcrowded (a) T = total power supply = N ( " T .42)] & 3 3 = [( ( = 0.67)x ( (2)x ( (4.

hence P(R < 0) = % ( 0#2 ) 19. and " T = [32 + 12 + (2)(0.847 (hours) J J Hence P(TJ > 10 hours) = 1 – P( T J # ! TJ "T J \$ 10 # 9 ) 3.4 (a) Let TJ be John’s travel time in (hours). R is normal with !R = !TB – !TJ + 1 = 10 – 9 + 1 = 2.8)1/2 = 19. now let R ' TB – TJ + 1. one should take the lower route to minimize expected travel time from A to D. TJ = T3 + T4 with ! T = 5 + 4 = 9 (hours).847 = 1 – %( 0. and B " T = [22 + 12]1/2 = 5 (hours) B Hence P(TJ – TB > 1) = P(TB – TJ + 1 < 0).4.603 & 0.449) & 0.8)(3)(1)]1/2 = 3. TB = T1 + T2 with !T = 6 + 4 = 10 (hours). "R = [ " TB 2 + " TJ 2]1/2 = (5 + 14.8 .26) = 1 – 0.8 = %(-0. .397 (b) Let TB be Bob’s travel time in (hours).327 (c) Since the lower route (A-C-D) has a smaller expected travel time of ! TJ = 9 hours as compared to the upper (with expected travel time = !TB = 10 hours).

6526 % 0.51 = 1 – &(0.3922) = 1 – 0.2"5 + 0. we have C = 20 – A – B Hence Thus S = 0.5 " 0.2 A + 0. S itself is normal. and 0.12 % 0.346 (cm) #S = Hence P(S > 4cm) = 1 – &( 4 ' 3.12 = 1 – &(0.1 B + 2.2 B + 0. !S = 0. As a linear combination of three normal variables.5 (a) To calculate probability.12 2 2 \$ 2 " 0.4.347 P(S > 4cm) = 1 –& ( (b) Now that we have a constraint A + B + C = 20m.1 " 1 " 2 = 0.3 A + 0.3 212 \$ 0. we first need to have the PDF of S. these variables are no longer all independent.1"8 + 2 = 3.282 .1(20 – A – B) ( S = 0.3"5 + 0.8 ) 0.51 (cm) Hence 4 ' 3 . for example.2 2 2 2 \$ 0.5.8 ) 0.8 (cm) as before.2 212 \$ 0. with parameters !S = 0.2"8 + 0.1212 % 0.8 (cm) #S = 0. with )AB = 0.577) % 0.1"7 = 3.

5)(3)(2)] = (36 + 4 + 12)1/2 = 52 Q \$ ! Q 40 \$ 62 ) Hence P(Q < 40) = P( % #Q 52 = &(-3.6 Q = 4A + B + 2C = 4A + B +2(30 – A – B) = 2A –B + 60.4. hence !Q = 2"5 – 8 + 60 = 62.00114 .0508) ' 0. 2 2 1/2 #Q = [4"3 + 1"2 + 2(2)(-1)(-0.

is the sum of them.0051 1 – )( 0 300 Q * 60 ' )( 0 ) = 0. 10) The annual max. P(city experiences (any) flooding) = 1 – P(city experiences no flooding at all) = 1 – P(X = 0) = 1 – (1 – p)10 = 1 – (1 – 0.0105. and \$ Q2 # \$ Q21 " \$ Q2 2 + 2 % Q1Q2 \$ Q1 \$ Q2 2 2 = 10 + 10 + 2&0. the total number of flood years over a 10-year period. p = P(Q > 100) = 1 – )( 100 * 60 ) 300 = 1 – )(2. Q * 60 ) = 0. 10).05 ' p = 1 – (1 – 0.57 300 = 104. hence ! Q # ! Q1 " ! Q2 = 35 + 25 = 60 (m3/sec). which translates into a condition on the design channel capacity Q0.5&10&10 = 300 (m3/sec). Q2 ~ N(25. Q. using the flooding probability expression from part (c): 1 – (1 – p)10 = 0. Q = Q1 + Q2. 2 = 0.05)1/10 = 0. following what’s done in (b).0051.7 (a) Let Q1 and Q2 be the annual maximum flood peak in rivers 1 and 2. we adopt a binomial model for X.9895 = 0.0105 ( 95 years. respectively.0105)10 = 1 – 0.4.32 (m3/sec) (b) The annual risk of flooding. ' \$ Q # 300 ( 17. (c) Since the yearly risk of flooding is p = 0.0105 (probability each year) Hence the return period is + = = 1 p 1 = 95.9949 300 ' Q0 = 60 + )-1(0.1 .59643882 0.9949) = 60 + 2. peak discharge passing through the city. and we have a course of n = 10 years.309) = 1 – 0.989510 ( 10% (d) The requirement on p is.5 300 . We have Q1 ~ N(35.

5 m3/sec will cut the risk by half.!Extending the channel capacity to about 104. .

8 P(T<30)=P(T<30|N=0)P(N=0)+P(T<30|N=1)P(N=1)+P(T<30|N=2)P(N=2) 30 # 35 30 # 40 30 # 30 % \$( ) ! 0 .217 .857) ! 0.2 " \$ (#0.5 " \$ (#1.5 ! 0.3 5 25 " 3 2 25 " 2 ! 3 2 % 0.3 =0.525) ! 0.5 " \$ ( ) ! 0 .2 " \$ ( ) ! 0 .4.

974 15.5 P(TA <60) = # ( 12.667) \$ # (0.952x2/3 = 0. 13.4 60 \$ 40 P(TB<60) = # ( ) " # (1.227 13.4) 60 \$ 70 ) " # (\$0.661 12 12 = " " " 0.65) 60 \$ 60 ) " # (0) " 0.417) 0.858 45 \$ 40 1 \$ # (0. 9 2 ! 6 2 ! 12 2 ) = N(100.9 N = total time for one round trip in normal traffic = N(30+20+40.746) " 0.4. 16.16) (a) P(on schedule in normal traffic) = P(N<20) = #( (b) 120 \$ 90 ) " # (1.52 TA = Normal time taken for passenger starting from A = T2 + T3 4 2 ! 12 2 ) = N(60.7 (d) P(45 ' TB ' 60) P(TB & 45) 60 \$ 40 45 \$ 40 #( ) \$ #( ) # (1.667) " 0. 15.417) 0.52) 9 2 ! 4 2 ! 12 2 ) R = total round trip time in rush hour traffic = N(30+30+40.952 12 P(TA<60) = # ( Percentage of passengers arriving in less than an hour = 0. = N(90. TB = N(40.933) " 0.65 = N(20+40.339 1 \$ #( ) 12 P(TB <60% TB >45) = .952 \$ 0. 12) 6 2 ! 12 2 ) = N(70. 12. (c) Under rush hour traffic TA = N(30+40.227x1/3 + 0.

216 ) " '( " 0.076) " ' ("1.5 " 0 0.5 " 0 0.7 ( 0 .718 .4.216 ) ! ' (1.10 (a) D=S1-S2 #D ! #S " #S ! 2 " 2 ! 0 1 2 Var ( D) ! Var ( S1 ) % Var ( S 2 ) " 2 & Var ( S1 )Var ( S 2 ) ! ( 0 .076) ! 0.5) ! ' ( 0.3 \$ 2 ) 2 ! .3 \$ 2 ) 2 " 2 \$ 0 .216 (b) P (".3 \$ 2 ) 2 % ( 0 .5 ( D ( .

2% 1% 2 " 3 2 ! 3 2 ! 2 \$ 0.37 (b) V1 " 60 X 1 V2 " 60 X 2 ' P(V2 ) V1 ( 400) " P( Z ( 400) ' # Z " 60 # 2 ) 60 #1 " 60 \$ 15 ) 60 \$ 10 " 300 % Z2 " 60 2 % 22 ! 60 2 % 12 ) 2 &1.95 5.05 ' P( X 5 + 70) " .37 *1. 2% 1% 2 " 3 2 ! 3 2 ! 2 \$ 0.839 2 n " 5 .645 " 8.95) " 5.95 1 70 ) 45 ) 3n .267 161 # 5 " #1 ! # 2 ! # 3 " 10 ! 15 ! 20 ! 3n " 45 ! 3n % 52 " % 12 ! % 22 ! 2 &1.37 0 2 25 ) 3n " 5.4 .8 ' % 5 " 5.37* )1 (. " .37 ! P( X 5 ( 70) " .6 \$ 3 \$ 3 " 28.8 ' % 5 " 5. 2 \$ 60 2 % 1% 2 " 25920 ' % Z " 161 ' P(V2 ) V1 ( 400) " 1 ) * ( (c) X 5 " X 1 ! X 2 ! X 3 400 ) 300 ) " 1 ) * (.6 \$ 3 \$ 3 " 28.621) " .4. */ .11 (a) X 5 " X 1 ! X 2 ! X 3 # 5 " #1 ! # 2 ! # 3 " 10 ! 15 ! 20 " 45 % 52 " % 12 ! % 22 ! 2 &1.

6 2 4 2 % 0.0478 =0.436 .4 \$ 15 & 18 ( Z & 0.8 \$ 0.4 \$15 & 18 * P( Z ) 20) & 1 " # ( (c) ( Z & 0.683 20 " 18 ) & 1 " 0.6 \$ 20 % 0.28 3.12 (a) P(X>20!Y>20)=P(X>20)+P(Y>20)-P(X>20)P(Y>20) & (1 " # ( (b) 20 " 20 20 " 15 20 " 20 20 " 15 )) % (1 " # ( )) " (1 " # ( )) \$ (1 " # ( )) 4 3 4 3 =0.2282 2.5+0.5 \$ 0.524 Z in normal distribution with ' Z & 0.4 2 3 2 & 2.683 ' Z & 0.6 \$ 20 % 0.436 * P( Z ) 20) & 1 " # ( 20 " 18 ) & 1 " 0.0478-0.4 2 3 2 % 2 \$ 0.7718 & 0.6 2 4 2 % 0.6 \$ 0.4 \$ 4 \$ 3 & 3.4.72 & 0.

4 4 " 0.13 (a) (b) (c) P(X!2)=1-P(X=0)-P(X=1) *5' *5' " 1 \$ (( %%0.6 # 0.046 T=H1+H2+B where H1.4 78. 40 2 .5 .8 # 40 # 40 " 78. T in Normal distribution with . 100 . 80 " 280 1 P(T 0 300) " 1 \$ / ( (d) .4 4 )0& )1& " 1 \$ 0. 40 2 .610.T " 100 . 80 " 280 . NB are the number of highway and building jobs won P(NH =1+ NB =0)=P(NH =1)P(NB =0) * 3' * 2' = (( %%0. 100 . H2 and B are the profits from the respective jobs.610.4. 20 2 .333) " 0. 20 2 " 60 300 \$ 280 ) " 1 \$ / (0.6 0 0. T " 40 2 .4 5 \$ 5 # 0.5 1 P(T 0 300) " 1 \$ / ( 300 \$ 280 ) " 1 \$ / (0.046 NH.T " 100 .255) " 0.4 2 )1& )0& =0.4 2 (( %%0.6 0 0. 2 # 0.2695 60 . T " 40 2 .4 5 \$ (( %%0.

4747 * P ( Z ) 0) ! 1 \$ ( ( ) ! 0.5.v=0.883 0.5 % 0. c.14 S: the amount of water available.1 # D2 ! 0.01 ! 0.4005 \$ (\$0.14842 ! \$0.01 ' D ! ln & D \$ 12 # D2 ! ln 1.4 2 ) ! 0. S follows LogNormal Distribution with E(S)=1.5 \$ 0.o.158 & Z ! ' D \$ ' S ! 0.07421) ! 0.4747 0 \$ 0.v=0.12 " 0.4.14842 ' S ! ln & S \$ 12 # S2 ! ln 1 \$ 0.4005 P(water shortage)=P(D>S) =P(D/S>1) =P((Z=ln(D/S))>0) Z follows Normal Distribution with: Var ( Z ) ! # 2 D " # S2 ! 0.4 # S2 ! ln(1 " 0.158 .5 % 0.12 ! 0.148 ! 0. c.o. E(D)=1.07421 D: the total demand of water.

596 0.887 ln 30 ' 3.15 P(Failure of antenna) = P(C<P) = P(C/P<1) Define B = C/D B also follows a lognormal distribution with !B = !C .4886 "C = 0.45 ) % 0.1 \$ 4 # 0.5xln(1+0.22) + (ln2.186 "P " \$ " \$ 4" % 0.7 ) % & ('1.029 0 .243) % 0.887 (b) P(P>30) = 1 ' &( (c) C is lognormally distributed with !C = ln90 – 0.8 – 0.138 2 = .135)(0.4886 – 3.186 ) % 1 ' & (0.865)4 = 0.9 Hence.1 \$ 4 # ln(1 \$ 0.693 = 3.135 (e) P(at least two out of 5 antenna failures) = 1 – P(X=0) – P(X=1) = 1 – 0.15 2 \$ 0.1844 2 C 2 R 2 V 2 2 2 2 = 0.2 \$ 0.152 = 4.568 + (-6.0058 P(antenna failure in 25 years) = 1-P(no damaging storm in 25 years) = 1-e-0.8655 – 5(0.5x0.15 (a) P is lognormally distributed with !P = !C + !R + 2!V – ln2 = (ln1.452)] – ln2 = 0. P(Failure of antenna) = P(B<1) = &( (d) ln 1 ' ! B "B ) % &( ' 1 .3x10-3 – 0.2 \$ 0.7 "B = " C2 \$ " P2 % 0.89) % 0.4.6953) – 0.887 2 % 0.029 = 0.08) + 2(4.5x0.!P = 4.12) + 2[ln120-0.186 = 1.5x0.0058x25 = 0.9 Mean rate of wind storm causing failure = 1/5 x 0.

6 &T # (c) 57.294) 2 # 2.717 ) # ( ("2.978 # 0.2 2 % 0.2 1 2 !C # ln 20 " (0.717 \$ Z # 0.16 (a) P(failure)=P(L>C)=P(C/L<1)=P(Z<1) Z in LN with ! Z # !C " ! L \$ Z # \$ C2 % \$ L2 in which \$ C # 0.6 # 0.356 ln 1 " 0.4.976 " 2.259 \$ L # ln(1 % & 2 ) # .014) # 1 " 0.022 0.2) 2 # 2.294 2 ' PF # ( ( # 0.259 # 0.356 (b) T=C1+C2 E(T)=E(C1)+E(C2)=20+20=40 Var (T ) # Var (C1 ) % Var (C 2 ) % 2 *) C1 ) C2 # 57.19 40 some other distribution .294 ! L # ln 10 " 12 (0.259 ' ! Z # 2.

95 ! 0.85 2 " 2 \$ 0.187 100 P(failure)=P(T<L)=P(T-L<0) 0 ( #Z ! )( ) %Z ! )( ( (100 ( 50) 18.0184 ! )( .197 % T ! 9.8 \$ 9.69 (c) ! 0.4.17 (a) C=F+B # C ! # F " # B ! 20 " 30 ! 50 % C ! (0.85 2 ! 18.3 \$ 50) 2 ( 50 ) 23.69 2 " (.85 T=C1+C2=2 # C =0.69 & ' T ! 18.3 \$ 30) 2 (b) ! 9.85 2 " 9.2 \$ 20) 2 " (0.

& F " (0.1) 2 \$ 16 2 " 0.6 & F " (0.4.48 & Z " (0.6 P(C’)= ( ( ii.3 \$ 0.48) 2 # (0.1 " 19.01 \$ 20) 2 P(C’)= ( ( 0 ' ('0.6 ' 20 " '0.779 0.1 " 19.52 " 0.3 \$ 0.233 0 ' ('0. 20 ' 19.12 P(F>20)=1.957 0.00043 0.18 16 (a) F " 18 # ! Ai i "1 % F " 18 # 16 \$ 0.( ( (b) i.233 F=18+16A % F " 18 # 16 \$ 0.4) ) " 0.12 P(no collapse)=P(C’)=P(F<M)=P(F-M<0)=P(Z<0) % Z " % F ' % M " 19.01 \$ 20) 2 " 0.52 .1) 2 \$ 16 " 0.4 & Z " (0.12) 2 # (0.6 ) " 0.4) ) " 0.

4 and a c.184)5 = 0.094 ) # % (\$0.094 "F = "a = 0.19 (a) a is lognormal with mean 0.184 0.4.204 = 4.382 "R = "C = 0.o.v. of 0.2 Hence R is lognormal with !R = ln 200 + !C = ln200 + ln0. of 25% W = 200 kips F = wa/g is also lognormal with !F = ln 200 + !a = 5.382 & 4.o.3g and c.2 P(failure) = P(R<F) = %( (b) \$ !R & !F " R2 & " F2 ) # %( \$ 4.32 P(none out of five tanks will fail) = (0.25 R = frictional resistance = WC Where C = coefficient of friction is lognormal with median 0.4 = 4.00021 .9) # 0.v.298 – 1.

4. according to CLT. and standard deviation "W = 50 #"T = 50 #5 min. = 250 min.129 . W is the sum of 50 i. with standard deviation "T = 5 minutes.i. random variables. W is approximately normal with mean !W = 50#!T = 50#5 min. hence. its average waiting time is !T = 5 minutes..13137085) & 0. Hence the probability P(W < 3.5 hrs) = P(W < 210 mins) W \$ ! W 210 \$ 250 % ) = P( "W 5 50 = P(Z < -1. Let W be the total waiting time of 50 cars.20 For any one car.d.

09 ) Now. Since each Mi is identically distributed as M in part (a). normal + normal).…. hence M is lognormal (because ln M is normal).2 )] = [ln(1.09)] .5 #\$x = ln %x – !x /2 = ln(10 / 1. we can calculate the mean and standard deviation of Mi as %i = exp(\$ + !2/2) = 2.04 ).354116378. hence P(M > 3) = P( ln M ' \$ M !M ( ln 3 ' ln(2 / 1.5 !F = [ln(1 + "F2)]0. let’s calculate the parameters of F and X: 2 0.5 0. ln M = ln F + ln X (i.04 & 1.5 !M = (!F2 + !X2) = [ln(1.5 = [ln (1 + 0.5 0.5 = [ln (1 + 0. Since M = FX.09 0.04 & 1.e. +i = %i [ exp(!2 ) – 1] 0. and 1.21 (a) First.000055 . where i = 1.999945364 * 0.04 & 1.5 = 0.4.731026675) T ' %T 120 ' 50 & 2 ( #P(T > 120) = P( ) +T 50 & 0.731026675 The total bending moment at A is T = M1 + M2 + … + M50. the sum of a large number of identically distributed.3 )] = [ln(1.04)] 2 0.322063427) * 0.2 / 1. similarly 2 0.5 0.09)] 2 0. independent RVs.093 (b) Let Mi be the moment at A due to the i-th force. the lognormal parameters of Mi are \$ = 0. let M denote the bending moment at A. From these.630447976 and ! = 0. 50 &0.2. with parameters 2 \$M = \$F + \$X = ln( ).5 # \$F = ln %F – !F /2 = ln(0.869116163) = 1 – 0.731026675 = 1 – )( 3.09) ) = 1 – )(1. hence we may apply the central limit theorem: T ~ N(50&2.09) ln(1.50.04&1.5 !X = [ln(1 + "X2)]0.

Y is approximately normally distributed with \$ 5000 #Y = #X = 15000 and \$Y = X % . covering an area of 20000 ! 16000 "1 = 0.007 (c) An individual’s probability of exceeding \$16000 is much higher than that for the mean of a group of people. while individual behavior can differ from the mean significantly. average value) becomes very centralized (i. almost constantly at \$15000) around the mean. 50 150 hence the desired probability is 16000 ! 15000 ) P(Y > 16000) = 1 –& ( 5000 / 150 = 1 –&(2.4. though they have the same mean.e. By the central limit theorem.45) ' 0. X has a uniform distribution between 10000 and 20000.4 20000 ! 10000 from x = 16000 to x = 20000. #X = (10000+20000)/2 = 15000.e. % 12 3 Let Y be the mean monthly salary of 50 assistant engineers. Uncertainty is reduced as sample size increases. (20000 ! 10000) 5000 \$X = . as X has a much larger standard deviation than Y. (b) Using properties of an uniform random variable. . and “collective behavior” (i.22 (a) Let X be the monthly salary (in dollars) of a randomly chosen assistant engineer.

T = V + D.4. 2 2 Var(V) = 100!Var(C) + 30!Var(T) = 100!2 + 30!5 = 1150 2 (kip ). The total vehicle weight (in kips) on the bridge is V = C1 + C2 + … +C100 + T1 + T2 + … + T30 Since V is a linear combination of independent random variables.80192694) # 0. (1150 + 120 ) ) = N( –100.23 (a) Let C denote car weight and T denote truck weight.v. 0. hence P(V > D) = P(V – D > 0) = P(S > 0) S % ( S 0 % (%100) & ) = P( 'S 15550 = 1 – \$(0. 15550 ).60385388) # 0. 2 1/2 0. Hence the c. 120) The difference S = V – D is again normal.0016 (c) (i) (ii) Let D denote the total dead load in kips. it mean and variance can be found as E(V) = 100!E(C) + 30!E(T) = 100!5 + 30!20 = 1100 (kips).5 15550 ). hence P(V + D > 2500) = P(T > 2500) T % ( T 2500 % 2300 & = P( ) 'T 15550 = 1 – \$(1.o. 0. where D ~ N(1200.948839) # 0.054 . T ~ N(1100 + 1200.031 (b) Assuming that the distribution of V approaches normal due to CLT.5 S ~ N(1100 – 1200.5 "V = 1150 / 1100 # 0. P(V > 1200) =1 – \$ ( 1200 % 1100 ) 1150 = 1 – \$(2.211 Let T denote the total (dead + vehicle) weight.

1 kg.581) = 1 – 0.24 (a) Let B be the total weight of one batch.1) = 1 –\$ (1.69% 101 % 100 (c) In this case. caused by the large standard deviation which makes the PDF occupy more area in the tails. we may apply the Central Limit Theorem: B is approximately normal with !B = 100 kg and #B = 40 " 0.158) = 1 – 0. hence 101 % 100 ) 40 " 0. .4.9431 & 5.5 kg = 100 kg (b) Since n > 30.7% P(penalty) = P(B > 101) = 1–\$ ( This penalty probability is much higher. Hence a large standard deviation (i.e. P(penalty) = P(B > 101) = 1 –\$ ( ) 40 " 1) = 1 –\$ (0. uncertainty) is undesirable.562816481 & 43. !B = 40"2.

187 = 8.44 ) FX ( y ) ! '( 0.294 ln y % 2.294 " = ln12 – 0.34 and 3.294 (b) 1 f X ( y) ! Hence By following the results of Examples 3.737 y 2 0.326 0.44 = 2.44 2 ] exp[% ( f Y20 ( y ) ! 20[' )] 0. # whose values are #= ln(1 \$ 0. the expected number of change of occupancy is 25.294) y ln y % 2.5x0. 3.25 Over the 50-year life.326 8.326 f Y20 ( y ) ! 8.9 9.326 ] y y 8.294 )n = e2. (a) The PDF of the lifetime maximum live load Yn or Y20 in this case is.57.9 . from Eq.3 2 ) = 0.294 0. Y20(y) will converge asymptotically to the Type I distribution with parameters un = 2.37.94( 2 ln 20 % and ( n ! ln ln 20 \$ ln 4& 2 2 ln 20 )+2.326 8.2942 = 2.294 2 2& (0.187 2 ln 20 ! 8. f Yn ( y ) ! f Y20 ( y ) ! 20[ Fx ( y )]19 f x ( y ) However X is lognormal with parameters ".44 1 ln y % 2.44 19 1 1 ln y % 2.44 2 ) ] exp[% ( 0.4.9 k = 8.9 ( ) exp[%( ) 8.

x FX ( x) " % \$ (\$z ) k #1 e #\$z !(k ) 0 dz The CDF of the largest value from a sample of size n is (Eq.2): x FYn ( y ) " [ FX ( y )] " [ % n \$ (\$z ) k #1 e #\$z !(k ) 0 dz ]n whereas. d 1 lim dx [ h ( x) ] " 0 x '& n The hazard function hn(x) is (Eq.28): hn ( x) " f X ( x) " 1 # FX ( x) \$ (\$x) k #1 e #\$x / !(k ) x 1 # %\$ (\$z ) k #1 e #\$z dz / !(k ) 0 \$ x k " k #1 x e !(k ) # % \$ z k 0 #\$x k #1 e x k #1 e #\$x " #\$z dz x \$ !(k ) # % z k #1 e #\$z dz #k 0 x d lim dx [ h x '& \$ # k !(k ) # % z k #1e #\$z dz 1 d ] = lim [ x ' & dx n ( x) 0 k #1 #\$x x e ] . 4. 4. then according Eq.3): n #1 fYn ( y ) " n[ FX ( y )] x f x ( y ) " n[ % \$ (\$z ) k #1 e #\$z !(k ) 0 (b) n #1 dz ] \$ (\$x) k #1 e #\$x !(k ) If the distribution of the large value is Type I asymptotic.26 (a) f X ( x) " \$ (\$x) k #1 e #\$x !(k ) Then. 4.77.4. 4. the PDF is (Eq.

the above limit becomes. using L’Hospital’s rule: # " (k " 1 " vx) ! ' ' "1 lim k " !x "! x \$# 0 The third term is of the form . thus. Therefore. . 0 !x k "1 e "!x '0 lim "!x ["!x k "1 (k " !x) ( (k " 1) x k " 2 (k " !x) " !x k "1 ] x \$# e = "1" lim since the order of the polynomial in the denominator is higher than that of the numerator. the distribution of the largest value converges to the Type I distribution.Using Leibnitz rule for the derivative of the integral. using the L’Hospital’s rule: x " 1 " lim " x k "1e "!x (k " 1 " !x) " ! [! " k %(k ) " & z k "1e "!z dz ] x\$# e "!x x k "1 (k " !x) " ! [! "k 0 x %(k ) " & z k "1 e "!z dz ] " ( k " 1 " !x ) 0 " lim "!x k "1 k " !x e x ( k " !x ) x \$# x \$# # The second term is of the form . Again. x lim " x k "1e "!x x k "1e "!x " [! " k %(k ) " & z k "1e "!z dz ]e "!x [(k " 1) x k " 2 " !x k "1 0 k "1 "!x 2 (x e x \$# ) x = "1" lim [! " k %(k ) " & z k "1e "!z dz ](k " 1 " !x) 0 x\$# x k e "!x Recall that # &z k "1 "!z e dz ' 0 %(k ) !k x "k k "1 "!z lim{[! %(k ) " & z e dz ](k " 1 " !x)} ' 0 x \$# 0 and k "!x lim ( x e )'0 x \$# Therefore.

2 )( L "18"(ln 27100) 3. ! = 1/3. the “design axle load” L can be obtained from 1 " exp["e " (1/ 3. 13550 and 33875 respectively.9 Similarly for period of 5.2)(80"18"(ln 27100 ) 3.e.8 (Y13550 = 0.2.999896 = 0.1 Hence. Hence mean maximum axle load is u Y1355 # 18 % [u n % *Y 1355 # ln 1355 0.o.33.4.2 ) ] = 1 – 0. Z follows an exponential distribution with mean = 3. for large n.2 )( L "50.2 )( y "18"(ln 27100 )3. 0.9 L = 57.577 & # % ] # 42.9 \$ n 1 / 3.v.2 ) ] Hence the probability that it will subject to an axle load of over 80 tone is 1 " FY27100 (80) # 1 " exp["e " (1/ 3.35 (b) +Y13550 = 50. exp["e " (1/ 3. the number of axle loads is 1355.39 42.2 *Y33875 = 16.7 ) ] # 0.2 ) 2 ) 2 (3. 10 and 25 years. are as follows: +Y6775 = 48 *Y6775 = 16.27 (a) Let Z = X-18.316 For a 20 year period.8 6 6\$ n2 '( Y1355 # 16. n = 27100 FYn ( y ) # exp("e "\$ n ( y "18"un ) ) # exp["e " (1/ 3.34 +Y33875 = 53. the largest value of an exponentially distributed random variable will approach the Type I distribution.2 1 / 3.3 *Y13550 = 16.8 and (Y33875 = 0.8 (Y6775 = 0.8 # 0.2) 2 # # 16. The corresponding mean values and c. the CDF of the largest value FYn ( y ) # exp("ne " !y ) which can be compared with the Type I distribution FYn ( y ) # exp("e "\$ n ( y "un ) ) By setting ne-!y = e "\$ n ( y " u n ) We obtain \$n = ! and =e un # "\$ n y \$ n u n e ln n ! For a 1 year period.2 Following the result of Example 3.2) ] .000104 (c) For an exceedance probability of 10%. i. Also.9 tons . the number of axle loads are 6775.

for the annual minimum DO. ! 1 # 0.125 (c) For Type I smallest value distribution ! 1 # u1 % 0.708mg / l . &2 "1 # 2 6' 1 Hence for monthly minimum DO level.5-0.4.18.212 "1 # # 0. ") follows a Type I Extreme Value distribution.577 '1 .0mg / l .112 % 5.87 2 # 0.792 % 6.577 # 1.112)] = 0. Because of the symmetry between the left tails of the normal distribution.87 (b) P[(Y1)30 < 0. 0.227 \$ 3 # 0.792)] = 0.21(0.5] = 1 – exp[-e6. the largest value for an initial variate of N(! .5 # 5.5) From E 4.577 # 0. ! 1 # 1. it also follows the Type I smallest value distribution with u1 # % 2 ln 365 \$ ln ln 365 \$ ln 4& 2 2 ln 365 \$ 3 # %3. 0.87 "1 # &2 6 ( 5.5 # 6.035mg / l .21 Similarly for the annual minimum level. it can be shown that the monthly minimum DO level also follows a Type I smallest value distribution with u1 # % 2 ln 30 \$ ln ln 30 \$ ln 4& 2 2 ln 30 ' 1 # 2 ln 30 / 0.112 Similarly.28 (a) Daily DO level = N(3.04 P[(Y1)365 < 0.06mg / l &2 6 ( 6.5] = 1 – exp[-e5.21 \$ 3 # 1. 0.5-1.792 ' 1 # 2 ln 365 / 0.87(0.435 \$ 1.

0285( D ' 79.45 \$ 100) 2 or 2 6# n & #n % % 0.4.45 \$100) 0. 1 ' FY ( D ) % 1 ' exp['e '0.061 P(damage to revised structure over 100 years) = 1 – e-(0.29 Let Y be the maximum wind velocity during a hurricane Y follows a Type I asymptotic distribution with !Y = 100 and "Y = 0.126 = 0.063 D = 175. the distribution parameters un and #n can be determined as follows: &2 % (0.172 .45 From Equations 3.75 0.00063 P(Damage to original structure over 100 years) = 1 – P(no damaging hurricane in 100 years) = 1 – exp(-0.0285 6 (0.00063 \$ 100) = 0.00063/2)(100) = 0.6 kph (c) Mean rate of damaging hurricane = ( = (1/200) \$ 0.75) ] = 0.61a and 3.75) ] % 0.0285(150 ' 79.0285 (a) P(Damage during a hurricane) = P(Y>150) = 1 ' FY (150) % 1 ' exp['e '0.61b.031 (d) P(at least one out of 3 structure with original design will be damaged over 100 years) = 1 – P(none of the 3 structure damaged) = 1 – (1-0.061)3 = 0.5772 u n % 100 ' % 79.126 (b) Let D be the revised design Hence.

*-/-%0*12%34+56*7.@I!e !# n & y ! u n ( H % 0*7=% % un " J +1 n ! +1 +1 n & +1 !% &\$% J J +1 n % # n " J +1 n K ' % D5>%'%#%-517=%@4>*52:%/84%1%L%M.@I!e !\$"#& S\$ ! !J". % J J +1 M.!"#\$% % &'(% )'*+.%U%\$"VSSK\$"#%L%!!%-@=% "%5E%Q1%L% % K& W# n ( " % K&\$"# W ( " !"JSS % '12%6"5"3"%5E%Q1%L%!"JSSK!!%L%\$"\$MS% .%-'. & +1 !% & !\$ " !J".\$(% <=4%-'.\$ " \$"# % % &B(% % &6(% % % <=4%-587%@>5B'B+4%0*12%34+56*7.%*8%9&!\$:%.*-/-%534>%#%-517=8%0*++%65134>?4%75%'%<.% N4164% un " J +1 M.G% % % % FYn & y ( " 4.( H % L%\$"\$\$\$J#% T4'1%5E%Q1%L%/1%U%(K!1%L%!J". # n " J +1 M.@4%A%+'>?487%3'+/4%2*87>*B/7*51%0=584%%%%%% C)D%*8%&E>5-%F.%-@=% P&)'-'?4(% L%P&Q1%R%S\$(% L% .'-@+4%!". ! 4.%2/>*1?%7=4%#O-517=%@4>*52%*8%!J". ! +1 +1 M. K .

4%:48/+78%5G7'*142%*1%L.%534:%'%#B-517.%-'.4:BA*CC477%2*87:*G/7*51E%0.% % S5:4534:E% k # > +1 N\$ T <">? # \$> % % &G(% U&)'-'94(% M%U&V1%W%X<(% M% \$ " 4.%*8%+5915:-'+%0*7.CJ"& y (k I % 0*7.%-4'1%!<%'12%6"5"3"%5=%>?@"%%A.4%-587%C:5G'G+4%3'+/4%!1%'12%8.*-/-%0*12%34+56*7.'-C+4%!">\$% D1%7.4% F*8.CJ<">?& > +1 N\$ " +1 +1 N\$ ( +1 !& ( ( #"P?Q % > > +1 N\$ %%%%%%M%4.%2/:*19%4:467*51%C4:*52%*8% ! n # 4.CJ!"\$QI%M%P?"!%-C.584%H)F%*8% % % !n FYn & y ( # 4.4% -'.*-/-%0*12%34+56*7.*8%6'84E%1%M%N\$% O% !%M%<">?% % % \$ > % # +1 !< " &<">?( > # #"P?Q % % R4164%7.C4%DDE%5:%7.%C4:*52%0*++%65134:94%75%'%A.%7.'C4%C':'-474:8%K% % % % '12% % % % ! n # eu k # \$n % +1 +1 n ( +1 !& un # ' & > +1 n " ((%% > > +1 n \$n # and n > +1 n ' % /8*19%7.!"#\$% % &'(% )'*+.4%-587%C:5G'G+4%0*12%34+56*7.CJ"& M%<"#P% % P?"! \$> ( I% X< .

v.516 % \$ V2 ( #L #V # f 1 (2)) ) % \$ D2 ( 3 # D (2 g ) 2 # D (2 g ) 2# L # f 2 2 1 2 2 & 100 & 0.8 2 ( ) % 0.152 ( 2 52 ( 1(2 & 32.2) 2 1 (2 & 32.0 fps c. #h " 100(9) 2 (0.3955+1.2) = 0.15 0.056 = 2.0158+355.673 ft .0052 ( ) % 0. Deviation 5 ft 0.01246 = 357.2) 1(2 & 32.25 0. \$ h " 18.02) ! 2.2) 1(2 & 32.9433+0.516 + 0.02 2 100 & 92 2 100 & 92 2 92 & 0. respectively.8 ( 2 1(2 & 32.9 ft (b) The second order mean of the hydraulic head loss 1 2 # h ! 2.02 1 2 100 & 9 & 0.02 9.2 std.32 h! LV 2 f 2 Dg Variable L D f V (a) Mean Value 100 ft 1 ft 0. 0.2) = 2.367 Hence.516 ft 2(1)(32.15 ( 3 (2)) = 2.2) 1 (2 & 32.05 0.02 2 ) ) % 1 .005 1.1006 + 0.2) #L # 2 2 #L # 2 2 # 2# # 2 #V2 # f 2 2 2 ) %\$D( 2 ) %\$ f ( ) % \$ V2 ( L ) \$ "\$ ( # D (2 g ) # (2 g ) # (2 g ) # (2 g ) 2 h V 2 L V D D V f D = 100 & 2 & 9 & 0.02 ) % 0. the first order mean and variance of the hydraulic head loss h are.o.516 % 1.4.8 fps Assuming the above random variables are statistically independent.15 ft 0.

25 0.8 \$ 10 #5 ( 750000 2 \$ 2 3 .05 0.000.2 0.8 \$ 10 #5 inch 3 3000000 \$ 72 \$ 144 4 L3 2 4 & P L3 2 4 & P L3 \$ 3 2 4 & P L3 2 2 2 2 2 2 ) ('E( 2 ) ( 'b ( ) ('h ( ) 'D % 'P( 3 3 2 3 4 &D % & E &b & h ( 2 \$ ) bh \$ ' b \$ ' h \$ = 100 2 ( & &b & h E 4& P L3 \$ 3 & E & b2 & h & E &b & h & E &b & h 4& P L3 \$ 3 & E & b & h4 4 \$ 500 \$ 180 3 2 4 \$ 500 \$ 180 3 2 4 \$ 180 3 2 2 2 ) ) 3 .2 2 \$ 12 4 \$ 500 \$ 180 3 ( ( ] [ (3e6) 3 \$ 72 \$ 144 3 3e6 \$ 72 3 \$ 144 3 3e6 \$ 72 \$ 144 5 2 = 1.925x10-12 = 4.000 psi 3. 2 \$ \$ 3e6 \$ 72 \$144 4 3e6 \$ 72 2 \$144 3 3e6 \$ 72 \$144 4 = 1.4.75x10-6 ft (b) The second order mean of the deflection 3 4 & P L3 (#3)(#4) 4 & P L3 (#1)(#2) 2 1 2 4 & P L (#1)(#2) 2 2 & D ! 1. the first order mean and variance of the deflection D are. 6 ( ) 750000 ( ( ( 3e6 \$ 72 2 \$ 144 3 (3e6) 2 \$ 72 \$ 144 3 3e6 \$ 72 \$ 144 3 ( 7.359x10-12 + 3.94x10-5 ft .v.o.8 \$ 10 ( [' E ( ) ('b ( ) (' h ( )] 2 & 3E & b & h3 & E & b3 & h3 & E & b & h5 #5 = 1.000 psi 72 inch 144 inch 4 PL3 Ebh 3 c.044x10-11 + 8.2 2 ( 4 \$ 500 \$ 180 3 \$ 3 2 4 \$ 500 \$180 3 \$ 3 4 \$ 500 \$ 180 3 \$ 3 ) ( 2 \$ 0 . respectively.176x10-13 + 7.33 D! PL3 . 8 \$ 3 . 0. 6 \$ 7 . 4 \$ 500 \$ 1803 ! 1. ' D % 6.05 std.6 inch 7. 3EI Variable P E b h I! bh 3 12 "D! Mean Value 500 lb 3. Deviation 100 lb 750.6 2 \$ 2 7.562x10-11 Hence.308x10-11 + 2.2 inch L = 15 ft = 180 inch (a) Assuming the above random variables are statistically independent.

of Z is 3.7.4430. 1 \$ "e " dx % 0.2)2 / 12 = 1/3.0) / (1 + 2) = 1.081 (b) Since Z = XY2W0. hence # x # 0 E(W) = 1/" ' 1. ) + (Y . we compute the approximate values E(Z) ' E(X) [E(Y)]2 [E(W)]0.5 ) Var (Y ) 0 0. E(Y) = 0 + (3 .5. Var ( X ) 0 . Also.5 & " % ln 2 .443". Var (W ) / + (X . ) 2 1 2 2 2 1 2 2 2 1 2 2 = Y 2W 0.443. and * (Z * (Z * (Z Var(Z) ' . i.5 ) Var ( X ) 0 2 XYW 0.7 ' 1. for W.5 & 1 # e " % 0. Y = 1.5 XY 2W #0.4.34 (a) Using Ang & Tang Table 5.e. E(X) = (4 + 2) / 2 = 3.o. 29.5 ' 3. hence Var(Z) ' 29.0)2/(1+2)2(1+2+1) = 1/2.5 ) Var (W ) where ) denotes "evaluated at the mean values X = 3. ) + (W /. since we know its median is 1.6 . / Var (Y ) 0 .v. W = 1. Var(W) = 1/"2 ' 2. Var(X) = (4 .51 Thus the c.6.5 = 3!12!1.1. Var(Y) = 1!2!(3 .

Y = (1 – L)X is a non-linear function of two random variables.012 ) 1 21 ( L 32X 5100. we have Y = (1 – 0.96 (100 4 0. Since X is log-normal.4. hence we need to estimate its mean and variance by E(Y) # Y(L = %L.20 P(acceptable) = P(Y ' 8) = P(Y’ ' ln 8) Y '( % Y ' ln 8 ( 1.585170186 When L is a constant (0.96 (0.64 . Var(Y) # * Var ( L) 1 * Var ( X ) .35 (a) Let X’ ! ln(X).e.95).96)(100) = 4. hence 4 .0001 + 0.589437912 = P( ) ' "Y' 0. we have the approximate parameter values \$Y # 1. X’ is normal with parameters "X’ # \$X = 0.L 50.05 + %X’ #1.2) 2 = 1000040.993 (b) (i) When L is also a random variable. Y’ is a constant plus a normal variate. i.95)X = 0. hence Y’ is normal with parameters Hence %Y’ = ln 0. / 0Y .20.450018146) # 0. 0L + % . and "Y’ = "X’ # 0.L 50.05 + X’. X = %X) = (1 – %L)(%X) = (1 – 0. and 2 2 / 0Y . 0X + % = 2( X 32X 5100.20 = )(2.00164400 = 1.05 X Taking the log of both sides and letting Y’ ! ln(Y) & Y’ = ln 0.64 (ii) Assuming a log-normal distribution for Y and using its estimated mean and variance from (i).589437912.202 / 2 = 4. and %X’ # ln 100 – \$X2 / 2 = ln 100 – 0.

312.312 / 2 = 1. 2 %Y \$ ln 4 – 0.338 P(Y & 8) = ' ( ) 0.312 = '(2.375) \$ 0.991 No.!Y = ln(1 # " Y2 ) \$ ln(1 # 1.338. this offer from the contractor does not yield a larger performance acceptance probability.64 / 16) = 0. . thus ln 8 ( 1.

67 .015-1*3..67E(S)0.67 S 0.463n 2.0050.665(ft3/sec)2 (b) The percentage of contribution of each random variable to the total uncertainty: *Qc *n ! "# 2 Var (n) / Var (Qc ) 0 74.2% 2 ) *Qc & ' *D \$ Var ( D) / Var (Qc ) 0 21.0.707 24 3 8 P(QC 7 30) 0 1 .67 D 1.5 # Var ( S ) =22.67 / 0.665 1 22. - 2 Var (n) + 0.36 (a) E(Qc)=0.2 - 2 # .1 S 0.0.463n .5=0.5S .116 # 41 5 0 ln # .4.02. 6 ( ln 30 .2% ( %# 2 ) *Qc & ' *S \$ Var ( S ) / Var (Qc ) 0 4.0 C 1 2 QC 0 22.463*0.5=41.116 .9958 0.5 # Var ( D) - 2 D 2.665 0 0 0.1 / 2. 0. 3.67*0.463D + .463E(n)-1E(D)2.0ft3/sec Var(Qc)= ! = *Qc *n "# 2 2 2 ) *Q & ) *Q & Var (n) + ' c \$ Var ( D) + ' c \$ Var ( S ) ( *D % # ( *S % # . 12 4 2 0 3.5 D .6% ( %# (c) QC follows the log-normal distribution with: # Q 0 41.707 ) 0 0.

2 ii. Var ( R) " 7 N ( )4Var ( R ) " 0. 60 & 50 ) " 1 & ' (1. / 1R .424 =1.48*0.5) * 7 N ( * ln( R * 1))4 # 1 8 R *1 5 2 2 2 =(ln2) (0.07 V " N R ln( R * 1) =1*1*ln(1+1) =ln2 =0. b) The function is highly nonlinear Check the second order term for the mean: 1 1 2V 1 9 ( R * 1) & R 1 6 . .375 * 2 2 2 1R 28 R *1 5 ( R * 1) which is not that small compared to the first order term.544 1.693 2 2 0 1V 0 1V Var(V)= .9213 " 0. 3 2 R 9 6 " R ln( R * 1) (0.07 ) P(T ( 60) " 1 & ' ( (b) i. + Var ( R) + Var ( N ) * . )the approximations are not good.793 ): V " 0.12+1.544 " 1.25+1.424*1 =0.4. / 1N .5) +(1/2+ln2) #1 =0.414) " 1 & 0.078 7.37 50 (a) T= ! Ri i "1 \$ T " 50 # 1 " 50 % T " 50 # 12 " 7.693 The approximation is not expected to be good because: 2 2 a) R and N have large variance.

968 . (b) #w = 0.25 10 2 2 400 "w \$ Hence.4.25 + 0. #w = 0.538 (c) The second order approximate mean of w is 1 2 '2w 1 2 '2w "w \$ 2 & # K ( 2 )" & # M ( )" 2 2 'K 'M 2 1 1 1 1 1 3 %3 / 2 " M %1 / 2 & # M2 (% )(% ) " K 1 / 2 " M % 5 / 2 = 2 & # K2 (% )( ) " K 2 2 2 2 2 2 1 3 = 2 % (200) 2 (400) % 3 / 2 (100) %1 / 2 & (20) 2 (400)1 / 2 (100) % 5 / 2 8 8 = 2 – 0. M = 100.25 & # M2 ( " K " M %3 / 2 ) 2 ! 0.29 Hence.25 & 20 2 ((20)(100) %3 / 2 ( )) 2 = 0.03 = 1. The first order mean and variance of w are 1 1 "k ! 400 ! 2 10 10 1 1 1 1 # w2 \$ # K2 ( " K % 2 ) 2 ! 200 2 ( ) 2 ! 0.38 w! (a) K M #K = 200 "K = 400.04 = 0.0625 + 0.5 M is also random with "M = 100 and #M = 20 The first order mean and variance of w are "w \$ "K 400 ! !2 100 "M 1 2 1 2 # w2 \$ 0.

3) *L "# 6.7 / .5.5 . # 1.5 %L "# )L "# 0.085 u "# runif( 100000 ! 0 ! 1) 1 W "# 0 - .57721566 .832 6 .1 MATHCAD statements: D "# rnorm( 100000 ! 4. "# )w "# 0.8 2 +L "# ln' *L( .ln2 ln2 77777 6 S "# ( D & L & W) n "# 100 j "# 0 88 n intj "# 0 & 20 .152 15 20 ' ( 2 ln 1 & %L . 0 # 3. 1 1 1 44 3 3 u 55 .)w 0 "# *w - 0.mean(S) *R # 21.0.102 *R "# 1.\$L )L *L \$L "# +L # 1.5. j n h "# hist( int! S) 4 1 810 h 5000 0 0 5 10 int mean( S) # 14.4 .2 ! 0.864 L "# rlnorm' 100000! +L ! \$L( *w "# 3.

!R "# 0.041 *\$R R "# rlnorm& 100000 .76 1 10 The result obtained with Mathcad with a sample size of 100.15 \$R "# & ' 2 ln 1 % !R (R "# ln& )R' + 1 2 \$R # 0.149 2 (R # 3. (R . \$R' .000 yields the probability of R<S is 0. ///// x "# [ ( R + S) .0] 99999 0 pF "# xi i# 1 100000 +3 pF # 9.009 .

b # 40 .15 ( b ' a) 2& q "# 2 b "# 2 %\$ therefore. 2 r "# q x1 "# rbeta( 10000 ! q ! r) F "# ( b ' a) %x1 ( a n "# 100 intj "# 0 ( 30 % j "# 0 )) n h "# hist( int! F) 1500 1000 h 500 0 0 10 20 int E "# rnorm( 10000 ! 1. 35000 30 j n therefore.6) C "# ++ * W %F E s1 "# C .6 ! 0.5 therefore.5.20) Distribution of F is beta \$ "# 20 a "# 0 & "# \$ %cov cov "# 0.2 MATHCAD statements: W "# rlnorm( 10000 ! ln( 2000) ! 0.125%1. & # 3 2 '1 q # 199.

000 is p1 " 0. the probability that the annual cost of operating the waste treatment plant will exceed \$35.3224 .9999 ! p1 #" s1i i" 1 10000 therefore.

1 2 * 2 'S 'S # 0.00 r "# 4. the result by MCS is very close to the exact solution.5.40 %S "# !S &\$S ) 'S "# ln 1 ( \$S +S "# ln) !S* . (b) S follows beta distribution MATHCAD statements: q "# 2.q .074 S "# rlnorm) 100000.398 +Y # .0 . D) 1 2 2 'D +D # 0. total demand of water is D MATHCAD statements: !S "# 1 \$S "# 0.3 (a) Define water supply as random variable S.'D* 000 / x "# ( S .5 \$D "# 0.0. a) ( a a # 0. therefore event of water shortage is Y<1 'Y "# 2 2 'S ( 'D +Y "# +S .0.884 100000 Define random variable Y=S/D.883 Therefore. D "# rlnorm) 100000 .385 2 +S # .+Y .252 b # 2.r) S "# v&( b .4 99999 1 pF "# xi i# 1 pF # 0.1 'D "# \$D +D "# ln) !D* .'S* !D "# 1. +D 'Y # 0.+D .475 pf "# plnorm) 1.'Y* pf # 0. q &%S & b "# a ( %S & q( r( 1 q &r q( r( 1 &( q ( r) q &r v "# rbeta( 100000 .00 a "# !S .+S .497 .

04 skew( pF) % /0.1 pF \$% . ln 1 + )S ." ### x \$% ( S ! D) 99999 & pF \$% xi i% 1 100000 pF % 0./ *D \$% )D 1 2 2 *D . S 0 rlnorm 1000 ( .D ( *D### " x 0 ( S 1 D) 999 & pFi 0 xj j% 1 1000 pF pF mean( pF) % 0.5 )D \$% 0.D \$% ln.D % 0.863 (C) MATHCAD statements: 'S is uniform distributed./ 1 2 - 2 )S \$% 0.80 ( 1. 1000 ( .40 *S % 0.4 for i 3 0 22 999 .1) *S \$% .903 n \$% 50 j \$% 1 22 n intj \$% 0 + 1.385 *S 2 'D \$% 1. 'S.S \$% ln.Si ( *S - D 0 rlnorm.0 4 h \$% hist( int( pF) j n Stdev( pF) % 0. 'S \$% runif( 1000 ( 0. 'D.322 .

5 int 1 .300 200 h 100 0 0 0.

2 %T2 "# 20 &T2 "# ln' %T2( ) \$T3 "# 0...644 T2 "# rlnorm' 100000! &T2 ! \$T2( T3 "# rlnorm' 100000! &T3 ! \$T3( TR1 "# T1 * T2 * T3 .669 ..2 %T2 "# 20+2 &T2 "# ln' %T2( ) TRA "# T2 * T3 1 2 2 \$T2 &T2 # 3..037 100000 (b) MATHCAD statements: ..0 +60 ..55 100000 (c) MATHCAD statements: \$T2 "# 0..x "# [ ( T2 * T3) 0 60] 99999 / pf "# xi i# 1 pf # 0...5. x "# ( TR1 ) 2. 0..3 1 2 2 \$T2 &T2 # 2.0) 99999 / pf "# xi i# 1 pf # 0. the one during rush hours is TR2 (a) MATHCAD statements: T1 "# rnorm( 100000 ! 30 ! 9) \$T2 "# 0..4 Define the total travel time for one round trip under normal traffic as TR1.976 %T3 "# 40 &T3 "# ln' %T3( ) 1 2 2 \$T3 &T3 # 3..

80.pfB. (d) The passenger left Town B at 2:00pm. % 0. the probability he will arrive on time is 0.776 Therefore.7% of passengers will arrive the shopping center during rush hours in less than one hour." ##### xA \$% [ ( TRA) ! 60] 99999 & pfA \$% xAi i% 1 100000 pfA % 0.55 pfB \$% plnorm* 60 ( 'T3 ( )T3+ pfB % 0.933 2 1 pfA.806 3 3 Therefore. .3 .T3 \$% 40 'T3 \$% ln* . therefore the traffic is normal P(T3<60|T3>45)=? MATHCAD statements: )T3 \$% 0.T3+ / 1 2 2 )T3 T3 \$% rlnorm* 100000( 'T3 ( )T3+ #### " x \$% ( T3 0 45) " ######## y \$% ( T3 0 45 1 T3 ! 60) 99999 pf \$% & yi & xi i% 1 99999 i% 1 pf % 0.776 .

773 .Verify the simulation result by directly calculate the probability by cumulative distribution function pf1 '( plnorm\$ 60 " !T3 " #T3% & plnorm\$ 45 " !T3 " #T3% pf2 '( 1 & plnorm\$ 45 " !T3 " #T3% pf3 '( pf1 pf2 pf3 ( 0.

5..x "# ( D 1 0.217 (b) the probability that D is less than 0...5 inch is .294 0.30 5S1 "# 6 7 2 ln 1 % 4S1 5S1 # 0.5) 99999 2 pF "# xi i# 1 pF # 0.77 0 10 var( D) # 0.3 \$2 0.3 \$2) \$ 1 & 0.7 % (2 % 0.3 \$2 * .5 S1 and S2 follow bivariate normal distribution (a) MATHCAD statements are: u1 "# runif( 100000 ! 0 ! 1) u2 "# runif( 100000 ! 0 ! 1) S1 "# qnorm( u1 ! 0 ! 1) \$0. D "# S1 & S2 n "# 40 j "# 0 // n intj "# 0 % 3 \$ j n h "# hist( int! D) 4 1 /10 5000 h 0 0 1 2 3 int &4 mean( D) # 3.858 100000 (c) S1 and S2 are jointly lognormally distributed 3S1 "# 2 4S1 "# 0.. \$( S1 & 2)+ .7 \$ .3 \$2 % 2 2 ' ) S2 "# qnorm( u2 ! 0 ! 1) \$( 0.

1
2
!S1 '( ln# "S1\$ & %S1
2

!S2 '( !S1

!S1 ( 0.65

%S2 '( %S1

S1 '( rlnorm# 100000) !S1 ) %S1\$
A '( log( S1)

,
.

uB '( -!S2 + 0.7 *

%S2
%S1

/
1

*# A & !S1\$0

2

sB '( %S2 * 1 & 0.7

B '( qnorm( u2 ) 0 ) 1) *sB + uB
S2 '( exp( B)

2
333
D '( S1 & S2

n '( 40

j '( 0 44 n

intj '( 0 + 3 *

h '( hist( int) D)
4

2 410

h

4

1 410

0

0

1

2

3

int

mean( D) ( 0.481
var( D) ( 0.321

The probability that D is less than 0.5 inch is:
3332
x '( ( D 5 0.5)
99999

6

pF '(

xi

i( 1
100000

pF ( 0.555

j
n

5.6
!F1 "# 35

\$F1 "# 10

'

ln( 1 &

%F1 "#

(
)

2
\$F1 *

%F1 # 0.28

2

!F1 +

,F1 "# ln- !F1. /

1
2

2

,F1 # 3.516

%F1

!F2 "# 25

\$F2 "# 10

'

ln( 1 &

%F2 "#

(
)

2
\$F2 *

%F2 # 0.385

2

!F2 +

,F2 "# ln- !F2. /

1
2

2

,F2 # 3.145

%F2

F1 "# rlnorm- 1000000 ,F1 0 %F1.
F2 "# rlnorm- 1000000 ,F2 0 %F2.
F "# F1 & F2
n "# 50

j "# 0 22 n

intj "# 0 & 150 1

j
n

4

1 210

h

5000

0

0

50

100
int

mean( F) # 59.96
stdev( F) # 14.14

150

h "# hist( int0 F)

(b) the annual risk that the city will experience flooding is the probability of flooding
###
"
x \$% ( F ! C)

C \$% 100

99999

&

PF \$%

RP \$%

xi

i% 1

PF % 0.011

100000

1

RP % 92.678

PF

The annual risk is 0.011, the corresponding return period is 92.7 years.
(c) If the desired risk is reduced to half of the current one, i.e., 0.0055, then the capacity should be
increased.
C \$%

for c + 100 ) 101 ** 300
###
"
x ' ( F ! c)
99999

&

p'

xi

i% 1

100000

c if p ! 0.0055
break if p ( 0.0055
c
C'c
C

C % 107

Verify the above result:

C \$% 107

"
###
x \$% ( F ! C)
99999

&

PF \$%

xi

i% 1
100000

,3

PF % 5.02 - 10

Therefore, the channel capacity should be increased to 107m3/s.

(d). Correlated case

!F1 "# 35

\$F1 "# 10

2
'
\$F1 *
(
%F1 "# ln 1 &
2
(
!F1 +
)

%F1 # 0.28

1
2
,F1 "# ln- !F1. / %F1
2

,F1 # 3.516

F1 "# rlnorm- 1000000 ,F1 0 %F1.
A1 "# ln( F1)

!F2 "# 25

\$F2 "# 10

2
'
\$F2 *
(
%F2 "# ln 1 &
2
(
!F2 +
)

1
2
,F2 "# ln- !F2. / %F2
2

%F2 # 0.385

,F2 # 3.145

1 "# 0.80
uB1 "# ,F2 & 1 2

%F2
%F1

sB1 "# %F2 2 1 / 1

2- A1 / ,F1.

2

u1 "# runif( 100000 0 0.0 0 1.0)
B1 "# qnorm( u1 0 0.0 0 1.0) 2sB1 & uB1
F2 "# exp( B1)
F "# F1 & F2
n "# 50

j "# 0 33 n

intj "# 0 & 150 2

j
n

h "# hist( int0 F)

4

1 !10

h

5000

0

0

50

100

150

int

mean( F) " 60.019

stdev( F) " 18.901

(d.b) the annual risk that the city will experience flooding is the probability of flooding
%%%
\$
x &" ( F # C)

C &" 100

99999

'

PF &"

RP &"

xi

i" 1

100000

1
PF

PF " 0.035
RP " 28.425

The annual risk is 0.035, the corresponding return period is 28.8 years.
Compare with the independent case, it shows the annual risk of flooding is increased due to the
correlation between the flow rates of the two rivers.
(d.c) If the desired risk is reduced to half of the current one, i.e., 0.0175, then the capacity should
be increased.
C &"

for c + 100 * 101 !! 300
%%%
\$
x ( ( F # c)
99999

'

p(

xi

i" 1

100000

c if p # 0.0175
break if p ) 0.0175
c
C(c
C

C " 110

Verify the above result:

C \$% 110

"
###
x \$% ( F ! C)
99999

&

PF \$%

xi

i% 1

100000

PF % 0.017

Therefore, the channel capacity should be increased to 110m3/s.

7 W and g are contant values W !" 200 g !" 32.0) k !" ( bk ' ak) %x & ak n !" 40 intj !" 0 & 1.1 00 * / / u 11 1 ln. The median value is equal to the mean value. ' n !" 40 . " 9. k.35 %% %0.934 0.3 %g% 6 . is a Type I extremal variate + * !" * " 37. #k !" 0..0 As q=r. is beta distributed q !" 3.08 q& r& 1 q %r ak !" #k ' q %\$k % q& r& 1 bk !" ak & \$k % q %r ak " 0. !" 0.0 r !" 3.0 & 15 % j n h !" hist( int( a) . then the distribution is symmetric. j !" 0 )) n intj !" 0.3 %g ' 0.0 ( 3.0 % j !" 0 )) n j n h !" hist( int( k) 4 1.577216 * .5 1 int Maximum ground acceleration.612 x !" rbeta( 100000 ( 3.2 Coefficient of friction.5.188 %( q & r) bk " 0. ln.5 )10 4 1 )10 h 5000 0 0 0.645 u !" runif( 100000 ( 0 ( 1) a !" . a.40 \$k !" 0.

119 (b) The resistance of the anchors for each tank is RA . 10 The probability that a tank will slide from its base support is 0.5 1 !10 h 4 5 !10 0 0 5 10 15 int (a) The resistance by friction R R #\$ k "W F #\$ W " a g n #\$ 40 intj #\$ 0.181 .0 % 100 " j #\$ 0 !! n j n h #\$ hist( int& F) 5 1 !10 h 4 5 !10 0 0 50 100 int ( ))) y #\$ ( R ' F) 99999 * pf #\$ yi i\$ 1 100000 +1 pf \$ 1.

0 + 0.3) break if pf * ( 0.( 3.( W ! 99999 ' pf ..( 32.40 2k .( ( R \$ F) F . 10 R # k !W " ra a g &&& % x # ( R \$ F) F # W! 99999 ' xi i( 1 pf # 100000 ra if pf ) ( 0.( 3.( xi i( 1 100000 /2 pf ( 3.08 .( 0.0 r .119 !0.0 g .2 RA Verify the above result: R .( 0. the error of 1a is ma q .( for ra .RA .1 .3) ra RA # ra RA ( 8.2 1k .119 !0.512 0 10 (c) Considering epistemic uncertainties: assume the error of 1k is mk.( k !W " RA a g &&& % x .

45) for i 6 0 55 999 ak " #k \$mki ' q \$%k \$ q& r& 1 q \$r q& r& 1 \$( q & r) q \$r bk " ak & %k \$ x " rbeta( 1000 ! 3.25) ma " rnorm( 1000 ! 1.Pf 74 mk " rnorm( 1000 ! 1. R " k \$W a g 222 1 y " ( R 0 F) F " W\$ 999 3 Pf i " j4 1 1000 continue Pf yj ..0 ! 3.0 ! 0. .u // ln. ln.0 ! 1.0 ! 0.0) a"* ' 1 ( + + 1 .35 \$% \$0.3 \$g) \$mai\$ 6 * " 0.577216 ( u " runif( 1000 ! 0.0) k " ( bk ' ak) \$x & ak (" ) ( 0..3 \$g\$mai ' 0.

295 stdev( Pf ) \$ 0.897 .n #\$ 40 j #\$ 0 %% n intj #\$ 0.897 The 90% value for a conservative probability of sliding is 0.876 v #\$ sort( Pf) Pf90 #\$ v900 v900 \$ 0.0 " 2 ! j n hpf #\$ hist( int& Pf ) 600 400 hpf 200 0 0 1 2 int mean( Pf) \$ 0.346 skew( Pf ) \$ 0.

976 %T2 "# %T1 T1 "# rlnorm' 100000+ )T1 + %T1( x1 "# log( T1) / ux2 "# .8 .198 2 %T1 )T1 # 2. 1 * 0.5 310 4 1 310 h 5000 0 0 20 40 int The external load is L !L "# 10 60 h "# hist( int+ T) .8 u2 "# runif( 100000 + 0 + 1) x2 "# qnorm( u2 + 0 + 1) .151 stdev( T) # 4.392 covT "# stdev( T) mean( T) n "# 40 covT # 0.sx2 & ux2 T2 "# exp( x2) T "# T1 & T2 mean( T) # 25. %T2 %T1 0 2 .' x1 * )T1(1 2 sx2 "# %T2. T2 !T1 "# 20 \$T1 "# 0.)T2 & 0.175 j "# 0 33 n intj "# 0 & 50 .8 Single pile capacity T1.20 ' %T1 "# ( 2 ln 1 & \$T1 )T1 "# ln' !T1( * 1 2 )T2 "# )T1 %T1 # 0. j n 4 1.5.

0026 . the probability of failure of this pile group is 0.!L "# 0.30 \$L "# & ' 2 ln 1 % !L (L "# ln& )L' * 1 2 \$L # 0. L) 99999 / pf "# yi *3 pf # 2.294 2 \$L (L # 2.. y "# ( T .259 L "# rlnorm& 100000+ (L + \$L' ..6 0 10 i# 1 100000 Therefore.

. 3 3 . 0 -. 1 1 2 2 1 1 2 .10 *u2 " 2 *# 2 " 25 *u22\$ 1 . 3 . 2 */10 *u2 " 2 *# 2 " 25 *u2 \$ 12 . 3 10 -. 1 2 . 1 1 . 3 10 -. 1 . *. 1 1 .9 Determine the solution to Problem 3. . -.5 ! y) . 1 2 */10 *u2 " 2 *# 2 " 25 *u2 \$ 2 / / u2 %& runif( 100000! 0 ! 1) y should be real value and greater than zero. 2 . . . *i*3 *. 2\$ 1 . . / 2 . 1 1 2 . 10 1 1 . 10 -.10 *u2 " 2 *# 2 " 25 *u22\$ 1 1 1 / 2 2 1 01 11 11 11 1 11 1 31 1 10 1 -. . */10 *u2 " 2 *# 2 " 25 *u2 \$ 12 " " *i*3 *. 1 1 2 . 4 .1 *. 3 1 . ./10 *u2 " 2 *# 2 " 25 *u2 \$ 12 " 2 . ./10 *u2 " 2 *# 2 " 25 *u2 \$ 12 " FYx( 0. .57 by MCS (a) f( x ! y) %& # \$ 6 2 x" y 5 1 ' 2 6 fX( x) %& ( f( x ! y) dy + *x " )0 5 5 (b) 2 x" y f( x ! y) simplify+ 3 * fYx( x ! y) %& 3 *x " 1 fX( x) 2 ' 3 *x " y FYx( x ! y) %& ( fYx( x ! y) dy simplify+ y* ( 3 *x " 1 ) 1 . -. 1 1 .10 *u2 " 2 *# 2 " 25 *u22\$ 1 " . *.1 2 2 simplify . # 2 */10 *u2 " 2 * 2 " 25 *u2 2 / .1 2 . . so 0 1 1 1 1 1 1 1 1 1 01 11 11 11 1 111 31 10 1 1 -. 2 . 2 2 1 1 2 . -.1 . 2 .10 *u2 " 2 *# 2 " 25 *u22\$ 1 1 1 / 2 22 . . . . 10 10 1 . . 1 2 . 1 10 . 1 . 1 2 .4 / 1 2 2 . 3 3 .Problem 5. u2 + 1 2 2 solve! y .

therefore the inverse function of FX is .- ni i.5 (c) 4 3 2 2 FX( x) .65 Verify the simultion result.5) 99999 2 p . 0.1 100000 p .5 3 y) dy 0 .0 .5 fX( x) dx 0 !x " !x 5 5 5 6 1( %& ) +1 1 2 & " !( 1 " 15 !u1) ) solve3 x & 3 3 ) FX ( x) + u1 0 & simplify 1) & +1 1 2) & 3 + 3 !( 1 " 15 !u1) ) ' * x should be greater than zero..2 3 1( %& ) 2 2 & 1 '10 !u2 " 2 !# 2 " 25 !u2 \$ )* + 2 IFYx .0.! 2 1( %& ) 2 &10 !u2 " 2 !# 2 " 25 !u22\$ ) ' * 1 3 1111111111 0 n .0 4 5 fYx( 0.( 1.. 1.. IFYx / IFYx .65000000000000000000 60.

1 2 2 . 1 1 1 1 . 1 3 10 -. . ./10 %u3 & 2 %+ 2 & 25 %u32. 1 \$. 11 1 1 2 2 . 12 & 2 %i%3 2 1 % FY ( y) \$ u3 * 1 1 simplify . . 1 1 2 . 1 . 12 \$ 2 . 1 3 10 -. 1 2 1 . therefore the inverse function of FY is u3 "# runif( 100000! 0 ! 1) . 1 -. . 1 . 1 . 1 3 3 10 10 -. 1 / 2 ./10 %u3 & 2 %+ 2 & 25 %u32. 1 . 1 % . 1 2 2 . 12 & 2 & i%3 2 %.10 %u3 & 2 %+ 2 & 25 %u32. \$1 % /10 %u3 & 2 %+ 2 & 25 %u3 . 2 \$ 2 & i%3 %/10 %u3 & 2 %+ 2 & 25 %u3 . .10 %u3 & 2 %+ 2 & 25 %u32. 1 1 2 . 1 / 2 / 2 Y should be greater than zero. 1 / 2 .10 %u3 & 2 %+ 2 & 25 %u32. 1 2 2 2 2 .4 1 . 1 1 2 . 1 . 1 3 3 10 10 -. 1 solve! y .u1 "# runif( 100000! 0 ! 1) 1 \$1 1 2 IFX "# & %( 1 & 15 %u1) 3 3 1 ' 3 6 2 fY( y) "# ( f( x ! y) dx * & %y )0 5 5 ' 2 3 3 FY( y) "# ( fY( y) dy * %y & %y ( 5 5 ) 2 0 . 1 2 2 . 1 3 10 -. 2 & 2 %i%3 1 1 . 1 . 1 1 2 . 1 3 10 -. 1 1 2 . 4 ./10 %u3 & 2 %+ 2 & 25 %u3 . 1 . .

003 Yes.2817 / .0.+0.- 0 stdev( IFYx) 3 2 stdev( IFY) 4 1+1 2 / . Check stdev(IFYx) . IFYx) .2828 stdev( IFY) .0.! 2 1( %& ) 2 &10 !u3 " 2 !# 2 " 25 !u32\$ ) ' * 1 3 corr(IFX.2 3 1( %& ) 2 2\$ ) & # 1 '10 !u3 " 2 ! 2 " 25 !u3 * + 2 IFY .09i .. there are statistical correlation between X and Y.0.

.5 ) X1 2 i 2 i ) X1 i ) u2 & i ) X1 i 3 0 0.35 0.15 ) X1 3 0.5 0.5.0) X1 "# for i ( 1 ! 2 '' 99999 X1 \$ 1 if u1 % 0..10 MathCAD statements u1 "# runif( 100000! 0.65 i i X1 \$ 2 otherwise i X1 u2 "# runif( 100000! 0..+ n "# ( X1 * 2 ) Y1 & 20) 99999 - PF "# n i i #1 100000 PF # 0.25 0.35 0.5 0.4 0..15 i ) X1 0 0.15 i i X1 \$ 3 if u1 & 0..0! 1.0) Y1 "# for i ( 1 ! 2 '' 99999 Y1 \$ 10 if u2 % i i Y1 \$ 20 if u2 & i i 1 3 1 3 ) X1 1 i 2 ) u2 % i Y1 \$ 30 if u2 * 1 ) X1 i i Y1 \$ 10 if u2 % i i Y1 \$ 20 if u2 % i i Y1 \$ 30 if u2 * i i Y1 \$ 10 if u2 % i i Y1 \$ 20 if u2 % i i Y1 \$ 30 if u2 * i i Y1 (a) .25 0.5 ) X1 ) u2 & 0.4 i 1 i 0.0! 1..35 1 2 i 0.2 0.35 3 ) X1 i 3 .

7 (c) """! n #\$ ( Y1 & 20) 99999 % PF4 #\$ n i i \$1 100000 PF4 \$ 0.704 PF1 The probability of the runoff Y is not less than 20 cfs when the precipitation is 2 inches is 0.7 P(Y>=20)=0.352 100000 PF2 PF3 \$ 0.5 100000 """""""! m #\$ ( X1 2 ' Y1 & 20) 99999 % PF2 #\$ PF3 #\$ m i i \$1 PF2 \$ 0.(b) """ ! n #\$ ( X1 2) 99999 % n i i \$1 PF1 #\$ PF1 \$ 0. X and Y are not independent (d) """! n #\$ ( Y1 10) .75 Therefore.752 as shown in (b) P(Y>=20|X=2)=0.

2 100000 PY( y ) #" PY & Py1 if y 10 PY & Py2 if y 20 PY & Py3 if y 30 The marginal PMF of runoff is plotted as follows.5 0 0 20 40 y .197 100000 %%%\$ n #" ( Y1 20) 99999 ! n i i "1 Py2 #" Py2 " 0. 1 PY( y )0.553 100000 %%%\$ n #" ( Y1 30) 99999 ! n i i "1 Py3 #" Py3 " 0.99999 ! n i i "1 Py1 #" Py1 " 0.

.

(e) #######" n \$% ( Y1 10 ! X1 2) ### " m \$% ( X1 2) 99999 & Py1 \$% n i i%1 Py1 % 0.506 99999 & m i i%1 #######" n \$% ( Y1 30 ! X1 2) ### " m \$% ( X1 2) 99999 & Py3 \$% n i i%1 Py3 % 0.296 99999 & m i i%1 #######" n \$% ( Y1 20 ! X1 2) ### " m \$% ( X1 2) 99999 & Py2 \$% n i i%1 Py2 % 0.198 99999 & m i i%1 PY( y ) \$% PY ' Py1 if y 10 PY ' Py2 if y 20 PY ' Py3 if y 30 .

678 Stdev( Y1) ! 7.5 0 0 20 40 y (f) Stdev ( X1) ! 0.519 Or directly by corr( X1# Y1) ! 0.The conditional PMF of runoff with precipitation = 2 inches is shown as follows.671 cvar( X1# Y1) ! 2. .701 " %! cvar ( X1# Y1) stdev ( X1) \$ stdev ( Y1) " ! 0.519 Therefore. the correlation coefficient between precipitation and runoff is 0.52. 1 PY( y )0.

-58K\$&/8\$-95\$ M&.\$\$ (\$% x 6\$-\$A"A#K\$. %&'\$ x (\$ i \$(\$ " % x (\$ 3+"*+ \$(\$!"*!32\$ \$ %4'\$ ./S\$95/M5\$-95\$/=>>\$9:.)!+\$ '\$ (\$%\$..7>5\$C5G&7/\$&MM5.A !"*!3 N ) & 1"1#*! \$ \$ J7-9\$ L\$ (\$ )6#\$ (\$ .A sN n & .)!+ 2 2 %8'\$ E\$ # x QA").\$ 8".L\$C5R5M-7./0\$ \$ F7-9.+)+\$J97M9\$70\$.A\$-.-95070\$70\$&MM5.J/\$0-&/8&C8\$85D7&-7..../0\$ B>-5C/&-7D5\$9:.+\$T\$1"22\$ '\$ (\$%\$./K\$-95\$50-7G&-58\$D&>=5\$.+\$-.A\$-.G\$ O&4>5\$ B"1"\$ \$ 1\$ &-\$ -95\$ +P\$ 07H/7L7M&/M5\$ >5D5>\$ -./\$(\$ # x \$ n ) % xi \$ .L\$-95\$-50-\$0-&-70-7M\$-.C5\$ -95\$ D&>=5\$ .-95070\$?B@\$ \$ #\$E\$.\$ \$ %x (\$%*.\$ 45\$ -!\$ (\$ 6#"./\$.-&4>5"\$ \$ \$ %M'\$ \$ E\$ # x QA")."L"\$ J5\$ .L\$ -95\$ -50-\$ 0-&-70-7M\$ 70\$ 1"1#*!\$ Q\$ 6 #".+ \$ ."3*K\$)#"+2'\$ .\$\$ (\$% x 6\$I\$A"A#\$ \$ \$ \$ \$ %x K\$ x T\$I\$A"A#\$ \$ \$ \$ \$ \$ '\$ n n !"*!3 !"*!3 K\$.+' 1 2!! ! 1 0 \$(\$ \$(\$ \$(\$3+"*+\$ n \$# .-95070\$?.=-0785\$-95\$C5H7.+\$U\$1".&M7-:\$./5607858\$9:.\$45\$ \$ t& x \$ .4-&7/\$ -95\$ MC7-7M&>\$ D&>=5\$ .=-\$&00=G7/H\$I/.+\$T\$1".-95070\$-50-\$ <=>>\$9:.+)+"\$ \$ O95C5L.+\$U\$1"22 2 2 (\$%*)"*3*K\$)A"1+2'\$ sx s K\$ x T\$-\$A"A#K\$.!"#\$ \$ !x " *!+ \$(\$.L\$-95\$.\$ x '\$ n n !"*!3 !"*!3 K\$.@\$\$ \$ #\$(\$.L\$ -\$ LC.

357) 2 0.99 = (65 – 2.745) 2 0. ++ . P( X J – X M > 2) = P(D > 2) D / \$D 2/0 = P( 0 ) !D 6/ 5 = 1 – 1(0.6.58 & 1 # n ' (2. n) i. and 2 2 . respectively. 6 2 ) n Hence (c) When n = 10. D ~ N(0. n) .6304 n # n = 240 # (240 – 50) = 190 additional vehicles must be observed. standard deviation = ++ n .58%6) = 239.6 * .58% 50 6 50 ) = (62. hence their difference n D = XJ – XM has an (approximate) normal distribution with mean = \$ – \$ = 0. n = 50. Both are approximately normal with mean \$ (unknown true mean speed) and 6 standard deviation .0092 6 / 50 .6 * 2 (( = 6 (( . 67.2 (a) Since x = 65.19) (in mph) 6 2 (b) Requiring 2. let X J and X M denote John and Mary’s sample means.81.e.005 50 # <\$>0. 65 – 2. P(D > 2) = 1–1 ( 2/0 ) = 1 – 1( 2. ! = 6. Before we proceed to (c) and (d).58% 6 .228 (d) When n = 100. a 2-sided 99% interval for the true mean is given by the limits 6 65 " k0.

x = 10000 cfs. n #1 = 1/3.") is fixed at 90%. x # t" / 2.376 0.9 = 1.350 0. t" / 2. where " = 0. n t0.7171 1.833 Hence.n #1 & 3000 n = 1000 t 0.7138 1. 11739 cfs] (b) Since the confidence level (1 .n #1 s n . the 2-sided 90% confidence interval for the mean annual maximum flow is [(10000 . n #1 ' 0. and increase the sample size until the half width is narrowed to the desired 1000cfs. To make it equal to 1000 cfs.05." = [ x # t" / 2.05.33333…. until t 0.1739.044) cfs] % [8261 cfs.1. .335 0.05.7056 t 0.05. n With reference to the following table. n which can be solved by trial and error. one must have t 0. We start with n = 20 (say).05. n #1 n 0.7081 1. while s is assumed to stay at approximately s 3000 cfs.367 0.7291 1.7207 1. n = 10.044) cfs.05.n #1 s n ].328 We see that a sample size of 27 will do. i. (10000 + 1739.6.7247 1. and t" / 2.387 0. n #1 can be considered as a n function of the sample size n only.7108 1.e. Sample size. the “half-width” of the confidence interval. s = 3000 cfs.n-1 20 21 22 23 24 25 26 27 1.358 0.3 (a) The formula to use is <!>1 .n #1 \$ t 0.342 0. hence an additional (27 – 10) = 17 years of observation will be required.

4 # 2.219850903 # 0.6. N = A/P 5 (b) N " !N i "1 1 2 3 4 5 1.881. SN = 0.048 5 5 \$1 (c) Substituting n = 5. and assuming N is normal. N = 1.61) # 0.154 ) 0.15392644 ! 0.02 ) n = 305.427) as a 95% confidence interval for the true mean of N. n – 1 N . hence an additional (305 – 5) = 300 piles should be tested.05 0.045 known.02 with 90% confidence would require k0.154.025.22 = +(–1.149 5 i S N2 ! (N i \$ N )2 2 # 1.15392644. = i "1 # 0.136 1. 1.4 (a) Pile Test No.070 1.0537 .645 0.507 0.776. (e) Since A = 15N.95 = N ! t0.907 1. we have (1. n where t0.219850903 into the formula S <%N>0.02 ) n * (1. (d) With & = 0.045 2 ) = 304.025.37 0.27293719) = (0.045 n ( 0. to estimate %N to '0. n – 1 = t0.025. P(pile failure) = P(A < 12) = P(15N < 12) = P(N < 12/15) =+( 12 / 15 \$ 1.

164274419 Hence the confidence level is 1 – & = 1 – 0.533 to 1. as t goes from 1. dof always indicates the tail area to the right of tp.1 = – 0.05 # 0.09) as a 90% confidence interval for the true mean of N.91.082137209 = 0.083472454'0. 4 = 2.082137209 ( & = 2'0.) t 0.o. we have (3672 ! 366. t&/2.746996. 4 = 1. t&/2.131846486. &/2) as decreasing linearly.f. hence 4 589778 4 <\$N>0.132). it 5 implies 589778 4 ) = 1.533.746996 – 1. t 0. Sx = ! (x i "1 i # x) 2 5 #1 = 589778 .6. from t-distribution table we have (at 4 d.1 to &/2 = 0. (and t&/2.213996232 = 0. &/2 should decrease from 0. Note: our convention is that the subscript p in tp.1 – 0. 4 = 300 / ( 5 We need to determine the corresponding &.e.6% .533 Hence.05.5 5 (a) Let x be the concrete strength. x " !x i "1 5 5 i = 3672. 4 increases as &/2 gets smaller in general) Hence we may use linear interpolation to get an approximate answer: over such a small “x” range (from t = 1.90 = 3672 ! t0. 5 where t0. 5 – 1 .1. we treat “y” (i.132 # 1.05.164274419 % 83. 4 589778 4 is only 300 (psi) wide.533 to t = 2. 4038.1 + m(1. with slope m= 0.083472454 2.132. On the other hand.09) = (3305.746996.533) = 0.4 % 2.05. dof (b) If the “half-width” of the confidence interval.

58 # . 15. x = 12.30) or 30 additional observations of truck weights would be required.005 3 30 ! n ) .6 (a) N = 30.5 + 2.826) = (9.5 – 2.5 tons Assume ! = 3 tons Assume ! = 3 tons < " x >0.k 0.674. Hence 2.5 – 2.9 Therefore.58 # or 3 \$1 n' n’ = (2.326) tons (b) Let n’ be the total number of observations required for estimating to ± 1 tone with 99% confidence. 85 +12.826. (60 . x + k 0.005 ! n = ( 12.58x3)2 = 59. 12. .6.5 – 2.99 = ( x .58 # 3 30 ) = (12.

the likelihood function of n observations on a Rayleigh distributinon is n L= 2hi \$! i #1 2 " hi 2 e! 2 Taking logarithm of both sides ln L = n ln 2 – 2n ln ! + % ln hi – 1 !2 % hi2 2 ( ln L 2n 2%hi #" ' #0 ! !3 (! ! From the given data.82 + …… 2. % hi2 = (1.52 + 2.836 n where & =& i #1 *2 ! # 2 * %hi %hi or ! # ) n n 2 . n = 10.6.32) = 80.7 fH(h) = 2h ! 2 "h2 e! 2 From Eq.42 * + ! = ) 2.4. 6.

s = 0.03 mg/l.0 mg/l .03 + 2.025.0196 <1.485 / 10 " 0. 2 at the 5% significance level to be "!#\$1.485 / 10 " 2. Therefore the value of the test statistic is 0.377) mg/l (c) < ! >0.0 mg/l ! > 2. (b) < ! >0. %& + t 0. 2.833.8331 =1.t 0.485 . 9 = (1.0 mg/l Null hypothesis Ho: Alternative hypothesis HA: Without assuming known standard deviation.683. 2.0196 with f = 10-1 = 9 d. 9 ) 10 10 0.t 0. 9 ' 0.o. hence the stream quality satisfy the EPA standard.833.025. we obtain the critical value of t from Table A.6.2.95 ' ' .485 mg/l One-side hypothesis test that the minimum concentration DO is 2.8 (a) ! = 2.485 = 2.749mg/l 10 10 .03 # 2 0. the estimated value of the test statistic to be "" !#2 0.05. ! = 2. which is outside the region of rejection.95 = %& .2622 ) = ( 2.f.2622 10 10 = ( %& .03 – 1.03 – 2.485 0.

73.2 " 124.2'#40o 25..97) ! 0.0' ) 2 ] ! 0.3 " 124. (d) & H 2 = Var ( h ) + Var ( L )(tan \$ )2 + Var( \$ ) x ( L 'sec2 \$ )2 = (0..01)2 + (0. " 40o 25.3 # 124.97) ft .3) 2 " (124.3 % tan 40o 25' ! 108.6'#40o 25..3 sec2 40o25’)2 = 0.135 ! ! 0.& h ! 0..98 0.98 or P (#2.4 ! 124. 3 1 2 SL ! [(124. " (40o 25.135' 5 #1 s\$ 0.33) ! 0.9 124.3 ft.3) 2 ] ! 0. Standard error of the mean = L ! n 3 40o 24.77 x 10-5)2 x (124.33 ) So <H>0.3) 2 " (124.0026 & H ! 0..73 ( H ) 108..77 % 10 # 5 radian Standard error of the mean = n 5 (a) L ! (c) h ! 3 ft .051 ft. 108.0001 = 0.6'".2' (b) \$ ! ! 40o 25' 5 1 2 S\$ ! [(40o 24.0' ) 2 " .051 or P (108.164' ! 4.4 # 124.6.01 ft H = h + L tan \$ H ! h " L tan \$ ! 3 " 124..01 ! 0.2 # 124.0024+0.01 3 #1 s 0.85 ft..0577)2(tan 40o25’)2 + (4.0001+0. (e) P (# k * / 2 ) H #H &H ( k * / 2 ) ! 0.0577 ft.98 108.98 = (108...85 # H ( 2.

n \$1 2 ) -----.819cm2 (d) t' / 2. n \$1 sr1 n ! 0.4 ! 2.5) 2 " (2.5) 2 # 2] ! 0.9467 and n = 17.0447)2 (2% x 2.1cm .9208 and n = 17.4 " 1.57cm2 (c) & A 2 = S r 1 2 (2% r 1)2 + S r 2 2 (-2% r2 )2 &A = (0.7 Assume n=10.5cm 5 1. t'/2.0447 (b) A = % ( r 12 – r 22) = % (2.01 / 5 ! 0.01 / 5 ! 0.4 " 2. t'/2.5)2 = 0.52 – 1.9 = 3.671 = 0.4 r2 ! ! 1. n = 21.2498 so from Equation (1).52) = 12.6 " 1.7 For n = 17.4 For n = 18.5) 2 # 2 " (2.8982 and n = 17. S r 2 ! 0.5)2 + (0.16 = 2.10 2.14 From trial and error.0447 (a) r1 ! Similarly. n = ( t' / 2.17 = 2.55.15 = 2. n is found to be 17.5 \$ 2.6 " 2. . t'/2.6 " 1.5 " 1. For n = 16.07 S r1 ! 0.6 " 2.4 \$ 2.5 " 2.0447)2 (-2% x 1. t'/2. So additional measurements are to be made = 17-5 = 12.1 So.6 \$ 2.5cm 5 1 2 S r1 ! [(2.01 5 \$1 S r1 ! 0.6.(1) 0 . S r 1 ! 0.

9375 # sin 59.9583 ! 23334.5 # 119.6.131 .5 # 119.1394% (c) The mean area of the triangle is A! 1 ab sin " ! 0.15 \$ A2 ! b sin "\$ a2 % a sin "\$ b2 % ab cos "\$ "2 %0.5 # 449.5 # cos59. " ! 59.89.11 (a) From the observation data we can get a ! 119. b ! 449.5 # 119.5 # sin 59. (b) sa2 ! 0.9583 # 0.9583 # 0.217m 2 (d) A 0. A % k0.5 # sin 59.73m 2 2 The variance of the area is: 1 1 1 2 2 2 ! 0.217. 23334.9 ! & A % k0.9375 # 449.95 !\$ A ' ! & 23334.131 % 0.64 # 7.73 % 1.57 ' .9375 # 449.217 ' ! & 23322.9375 .64 # 7.15 . sb2 ! 0.1394% ! 52.23346.73 ( 1. s"2 ! 0.9583 # 0.5 .9583 .086m 4 \$ A ! 7.05 !\$ A .

025 !# A .06 " 0.52 .52 ! 4824.12 .21. sb2 ! 0.06 " 120.696m # A2 ! b# a2 \$ a# b2 ! 0.06 .277 ! 22. 4824.4833.3\$ and Var (C ) ! E A (Var (C | A) ) \$ VarA ( E (C | A)) ! 200002 \$ VarA (10000 \$ 15 A) ! 200002 \$ 152 " # A2 ! 200002 \$ 152 " 22.96 " 4.056 ! 400004962.52 " 0. b ! 120.12\$ P(C * 90000) ! 1 ' P(C + 90000) ! 1 ' .182 \$ 0.62 & (d) It can be shown that E (C ) ! E A ( E (C | A)) ! 10000 \$ 15 " E ( A) ! 10000 \$ 15 " 4824.5 " 120.182 .696 & ! % 4815.6.3 ) ! 0.42 ' 1.42m 2 2 (b) 1 1 2 2 2 # A ! 4.12 (a) From the observation data a ! 80. sa2 ! 0.5 " 80.95 ! % A \$ k0. A \$ k0.6\$2 # C ! 20000.96 " 4.42 \$ 1.277 Thus A ! 1 ab ! 0.696.056m 4 (c) A 0.35 20000.975 !# A & ! % 4824. ( 90000 ' 82366.5 " 80.42 ! 82366.

06 / 10 4.13 Let x denotes the mileage (a) ! ! 37.o. we obtain the critical value of t from Table A.87 4.06 (b) One-sided hypothesis test that the stated mileage is smaller than 35 mpg: Null hypothesis Ho: " = 35 mpg Alternative hypothesis HA: " < 35 mpg Without assuming known standard deviation.37.06 ' & ! ) 37.6. Therefore the value of the test statistic is 1. which is outside the region of rejection.448.40.262 * 10 10 .50.30 % .4 # 2.448. (c) ! 0.4 ( 2. 2 at the 2% significance level to be #!\$%-2.262 .06 / 10 with f = 10-1 = 9 d. the estimated value of the test statistic to be #! 37.4 " ! ! 4.9 ! ! * 10 10 .87>-2.95 " " ' & ! ) ! # #0.025.4 # 35 ! # 35 ! ! 1. + 4. + ! \$ 34. hence the stated mileage is satisfied.2.06 4.9 ! ! .025.f. ! ( #0.

06 .9) = 0.9 c.7. = ln (81.5 / 76.v.o.1 Median = 76.

089 1.855 -0.9 73.810 0.3 72.166 0.225 -0.8 81.1 77.903 1.476 0.286 0.113 -0.3 70.238 0.5 74.2 78.735 0.302 3.190 0.298 0.048 0.3 80.581 0.302 1.905 0.036 0.667 0.9 76 76.762 0.1 76.6 72.2 P 0.870 2. %' ( x ( ' The standard variate is S & %# ( X % \$ ) ! With CDF of: FS ( s ) & exp % e % s " If x is observed with frequency of p .952 s -1.619 0.666 -0.143 0.4 78. the observed data can be arranged in the right table and plotted in the following figure.8 75.095 0. then its corresponding standard variate is s & % ln ! % ln( p ) " Thus.333 0.506 -0.857 0.020 .554 1.7. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 x 69.361 -0.094 0.2 78.429 0.8 85.524 0.571 0.436 0.714 0.8 75.2 (a) The CDF of Gumbel distribution is: FX ( x ) & exp ! % e %# ( x % \$ ) " .381 0.4 79.4 77.

1-%. . /. \$. -..617 0. )C:?8:68"7:6<:C5"! 0.. +..860 20 ( ni # ei ) 2 ( ni # ei ) 2 ei 0.723 S! (c) Perform Chi-Square test for Gumbel distribution Interval 65-70 70-75 75-80 80-85 & Observed frequency ni Theoretical frequency ei 1 5 11 3 20 0. Comparing with the & ( ni # ei ) 2 ei calculated.723 which gives X # 74. \$ ! 0... For a significance level \$ ! 2% .295 .240 2.007 2.356 7..723) 3.415 6. (. (%.. -%.41 . 23456758"9:.98.. .3942 S " 74. !"#"\$%\$&'()"*"+'%+(\$ '. . c.986 The degree of freedom for the Gumbel distribution is f=4-1-2=1... the Gumbel distribution is a valid model for the maximum wind velocity at the significance level of \$ ! 2% .<9=9"><?8"75@AB<CD"E (b) The linear regression of the data gives the trend line equation of X ! 3.924 0.1 ! 5..891 0. '%.602 8.%. 1(%.295( X # 74.164 0.. % ! 72.. .&. \$%..020 1.3942 Thus..723 ! 0.770 7.

939 3.3125 0.1 0.7.9375 From Figs.876 Perform Chi-square test for normal distribution: Observed Theoretical frequency ni frequency ei 15.143 0.8125 0.5 3.943 ( ni ( ei ) 2 1.000 3.5625 0.3 17.317 2 ( ni ( ei ) 2 ei 0.1 19.6875 0.5 19.7500 0.533 0.4 18.730 4.3750 0.581 0.401 0.5-16.0742 \$ 2.890 ( ni ( ei ) 2 ei 0.5 16.000 4.3 (a) m "! m N #1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16.6 18.000 3.6 16. 7. it can be observed that both models (normal and lognormal) fit the data pretty well.0 16.1 18.5 * 3.8 17.037 0.5-17.000 1.5 ) 0.5-17.166 15 Perform Chi-square test for lognormal distribution: Observed Theoretical Interval frequency ni frequency ei 15.3b.021 0.8) ( 0.989 1.5000 0.126 0.6250 0.342 0.8 19.8 sU2 \$ 1.000 15 1.312 & \$ ln(1 # %2 1.82 1 2 ' \$ ln ! ( & 2 \$ ln(17.024 1.5 16.1875 0.312 2 ) ln(1 ) \$ 0.1 16.000 4.028 ( ni ( ei ) 1.5 17.4375 0. (b) U \$ 17.0 17.000 1.5-20.3a and 7.226 .5-19.0625 0.8 17.5-16.816 3.163 1.4 20.8750 0.9 18.1250 0.772 0.2500 0.074 \$ # !2 17.5 18.5-18.404 2.

000 4.064 0.028 0.5 ! 4.034 7 17.366 0.8 0. c.889 0.023 0.083 5 17.059 0.493 0.839 0.105 4 16.052 6 17.010 0.533 0.4 0. ( n \$ ei ) 2 As the ! i in Normal distribution is less than that of the Lognormal distribution.133 15 0.200 0.667 0.688 0. (c) Perform the K-S test as follows: F D F D (Normal) (Normal) (Lognormal) (Lognormal) k x S 1 15.783 0.046 max ! ! ! 0.5-20.933 0.5 19.400 0.003 0.010 0.840 0.000 0.067 0.886 0.184 0.008 2 16.2 # 7.800 0.003 0.733 0.018 0.3 0.8 0.005 12 18.081 0.022 11 18.004 0.530 0.051 8 17.014 9 18.960 0.778 1.537 0.5 18.085 0.5-18.0 0.1 0.8 0.052 3 16.9 0. ei the 2 distributions are both valid for the rainfall intensity at the significance level of " # 2% .316 2.100 1.098 0.6 0.102 ! 0.352 0. the ei For a significance level " # 2% .1 0. ( ni \$ ei ) 2 calculated.500 0.867 0. Comparing with the ! Normal distribution is superior to the Lognormal distribution in this problem.777 0.047 15 20.467 0.676 0.095 0.033 0.5-19.282 0.048 0.99 .4 0.048 0.000 15 4.590 0.1 1.102 0.016 1.267 0.271 0.026 14 19.105 The K-S test shows that the Normal distribution is more suitable in this case as it gives smaller maximum discrepancy.600 0.954 0.023 0.729 0.382 The degree of freedom for the Normal and Lognormal distribution are both f=5-1-2=2. .547 0.062 0.180 0.006 10 18.045 0.98.606 0.086 0.0 0.133 0.738 0.040 0.017 13 19.17.333 0.517 0.000 1.6 0.1 0.

7.Fig.3a .

7.3b .Fig.

4 (a) .7.

19 st2 ! 359937.669 0.29 29.158 33 862.250 0.825 0.300 0. As the maximum discrepancy is smaller than the critical value.636 0.38 0.143 28 678.182 36 1057.34 3.82 14. c0.004 0.218 0. (c) Time-to-failure (hours) 0 – 400 400 – 800 800 – 1200 > 1200 # Observed Frequency ni 20 7 7 6 40 Theoritical Frequency ei 17.150 0.988 0. Mean time to failure = 1 " ! 670 hrs.93 0.071 23 633.075 0.001 0.114 0.950 0.171 35 952.375 0.182 ! 541.99 (ni & ei ) 2 ! 1.500 F(Normal) D(Normal) k x S F(Normal) D(Normal) 0.005 0.000 0.180 32 861.600 0.112 25 656.740 0.425 0.043 0.000 0.615 0.525 0.362 0.75 39.45 0.025 0.58 380 393.012 0.850 0. (d) Perform K-S test as follows: k 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 max Where t x 0.079 0.120 0.40 0.08 8. 2 = 5.098 0.22 0.55 8.800 0.045 0.750 0.761 0.584 % 5.067 1. . 0.446 0.97 0.040 0.95.161 30 813 0.726 0.034 0.625 0.400 0.58 124.95 0.021 0.070 0.13 0.627 0.975 0.200 0.061 0.040 39 1885.147 37 1296.104 0.628 0.725 0.727 0.13 0.021 0.880 0.507 0.170 0.47 1.705 0.441 0.09 151.227 0.(b) The minimum time-to-failure = 0 hr.275 0.013 0.204 0.044 0.550 0.04 0.41 2.180 34 895.575 0.025 21 412.01 0.89 0.278 0.022 0. At the 5% significance level.051 0.063 0.584 x2 distribution with f = 4 .05 D40 ! 0.119 0.21 .93 0.67 40 (ni-ei)2 (ni-ei)2 / ei 4.039 0.067 0.98 9. the exponential distribution is a valid distribution model.054 40 2959.44 163.110 0.156 31 855.057 0.93 0.462 0.98 0.775 0.125 0.28 121.09 84.450 0.650 0.796 0.074 0.849 0.87 0.435 0.65 0.78 2.900 0.85 62.621 0.002 0.131 27 672.875 0.52 0.95 216.57 0.175 0.99 ' ei i !1 4 In this case Hence.167 0.45 6.049 22 470.90 5.925 0.225 0.475 0.325 0.089 0.095 24 646.4 298.09 85.056 0.146 29 735.100 0.088 38 1407.2 The sample size is 40.129 26 658.050 0.700 0.639 0.675 0.2 = 2 degrees of freedom and \$ = 5% significance level.37 S 0.858 0.941 0.724 0.1 54.03 0.350 0. the exponential distribution is verified.8 0.

08 .0001 0.6084 0. Nos. Observed ni 0 6 0 1 8 8 6 2x3 + 3x2 "2 + 4x1 = 16 #20 #24 \$! Theoretical Probability (v = 1. of Observed of veh.76 #1.0854 x2 distribution with f = 3-2 = 1 degree of freedom and % = 1% significance level.0843 0. sv2 ! 1.361 0. the Poisson distribution is a valid distribution model. c0.635 ( ei i !1 3 In this case Hence.301 0.0 #20.5 (a) From the observation data we can get v ! 1.635 (n i ' ei ) 2 ! 0. veh/min Freq.7.338 Theoretical Frequency ei 6. #0.99.22 6.243 (b) Total no.1 = 6.2 20 (ni-ei)2 (ni-ei)2 / ei 0.0 24 ! 1.0004 0.02 7.1698 .2) 0.5776 0.1698 & 6.

20 0.70 0.94 0.95 0.84 1.955 0.15 FS(s) 0.79 1. The slope of a straight line on this paper is. s<0 1 # s2 2 s 0.965 0.90 3.30 0.59 2.10 0.65 2.985 0.60 0.80 0.75 0.65 0.72 2.26 1.99 0. s!0 .55 0.93 0.01 1.975 0.45 1.35 1.6 (a) f X ( x) " S" x \$2 e 1 x # ( )2 2 \$ !0 X \$ dX "\$ dS X " S\$ .95 2.37 2.49 2.97 0.6.50 0.46 0.26 FS(s) 0.31 2.7.96 0. 7.995 The corresponding probability paper is shown in Fig.80 2.98 0. .25 2.03 3.18 1.40 0.55 1.54 2.67 0. So f S ( s ) " f X ( s\$ ) = s %e =0 FS ( s ) " 1 # e 1 # s2 2 1 # s2 dX " s %e 2 dS .90 s 2.92 0.45 2.67 1. dX "\$ ds which is the model value of X.85 0.

7. the slope of the fitted line will give the most probable stress range. However.6 (c) On the basis of the observed data. . if the observed data indeed fell on a straight line.Fig. Rayleigh distribution does not appear to be a suitable model for live-load stress range in highway bridges.

579 0.046 0.25 10.125 0.4336 10.025 0.873 1.113 0. c0.075 0.200 0.174 0.500 4.8836 2.991 (n i ! ei ) 2 In this case & # 1.010 0.022 0.06 24.250 " ! ( Ki ! K ) 2 ni ( Ki ! K ) 2 0.672 % 5.075 0.7.0046 ! (ni-ei)2 (ni-ei)2 / ei 0.825 2.250 > 0.025 3.672 x2 distribution with f = 5-3 = 2 degree of freedom and \$ = 5% significance level.8281 0.375 sK2 # 0.010 0.002 0.100 – 0.04 83 (b) K (per day) < 0. .150 0.991 ei i #1 5 So the normal distribution is acceptable.375 3.000 0.023 0.003 0. of Ki observation( ni ) ni Ki 0.275 0.7 .09 19.022 0.650 14.95.107 0.175 0.048 0.225 0.039 0.000 0.200 – 0.025 0.325 " ! 1 11 20 23 15 11 2 83 ! K # 0.56 18.046 0.5625 8. No.(a) The middle point of each range is used to represent the range.150 – 0.2 = 5.7616 0.100 0.425 Observed frequency ni 12 20 23 15 13 83 Theoretical frequency ei 11.099 0.025 0.

8.806 0.224 0.316 0.364 .318 0.990 0.894 0.955 0.922 0.975 0.90 s 0.95 0. 0 % s % 1 S! =0 (b) .227 0.837 0.091 0.7.970 0.20 0.992 0.975 0.949 FS(s) 0.05 0.98 0. S > 1.987 0.447 0.965 0. S<0 s 0.70 0.93 0. X #a & x!a r X #a & x !r"a When s ! 1 ! r ' values of X at FS(0) and FS(1. (c) m 1 2 3 4 5 6 7 8 X 18 28 32 34 36 45 48 50 m N "1 0.97 0.982 0.96 0.65 0.80 0.985 0.548 0.85 0.40 0.0) correspond to a and r+a which are the minimum and When s ! 0 ! maximum values of X.980 0.985 0.30 0.60 0. X ! Sr " a.0 =0 . !r r dS dX 2 So f s ( s ) ! f x ( sr " a ) ! 2 ( sr ) \$ r dS r = 2s .995 FS(s) 0.742 0.045 0.94 0. elsewhere FS(s) = S2 .775 0.10 0.8 (a) X #a dX .964 0.50 0.92 0.75 0.866 0. 0 % s % 1 = 1 .55 0.707 0.273 0.977 0.959 0.182 0.99 The corresponding probability paper is shown in Fig 7.632 0.136 0.

909 0.818 0.500 0.455 0.636 0.682 0.864 0. 7.773 0.955 Fig.9 10 11 12 13 14 15 16 17 18 19 20 21 53 53 55 56 58 62 64 66 69 71 71 72 75 0.409 0.8 .727 0.591 0.545 0.

41 0.40 0.8571 0.1429 0.70 0.2143 0.7857 0.42 0.43 0.5000 0.2857 0.0714 0.6429 0.7143 0.96 m N !1 0.58 0.68 0.5714 0.49 0.35 0.4286 0.87 0.9 m 1 2 3 4 5 6 7 8 9 10 11 12 13 shear strength (S) (Ksf) 0.48 0.9286 .7.3571 0.75 0.

184 . of observation( ni ) niGi 0 6 34 132 179 218 183 146 69 30 3 0 1000 0 12 102 528 895 1308 1281 1168 621 300 33 0 6248 27.487 448.793 13.085 1.586 12.550 5.078 22.063 278.148 522.5 0 6 34 2.408 103.520 2.070 7.966 (Gi ! G ) 2 ! ni (Gi ! G ) 2 ( ni ! ei ) 2 ( ni ! ei ) 2 ei 6.688 37.062 0.993 ! Perform the Chi-square test for normal distribution Observed Theoretical Interval frequency ni frequency ei 0.566 3.086 0.000 108.054 1.248 ! sG2 # 2.542 18.5 1.5 2.574 14.273 358.582 33.5-1.325 67.572 422.000 2990. (sec) (b) The middle point of each range is used to calculated the sample mean and sample variance as follows: Gi ! 1 2 3 4 5 6 7 8 9 10 11 12 " No.368 48.113 40.5-3.496 G # 6.046 10.7.10 (a) Number of Observatio 250 200 150 100 50 0 1 2 3 4 5 6 7 8 9 10 11 12 Acceptance Gap G.5-2.745 0.683 667.586 3.558 0.

5-3.125 668.5-7.5-11.5 ' 2.012 0.5-6.204<97.248) " 0.5 10.537 47.452 138. of Observations ni Gi * ni ! ni (Gi-\$G)2 (Gi-\$G)2 ni / +ni (x10-3) 0 6 34 132 179 0 0.5-12.417 10.204 Perform the Chi-square test for lognormal distribution Observed Theoretical Interval frequency ni frequency ei 0. the normal distribution is more suitable for this problem.5 2.5 9.067 1.621 168.5 8.688 .577 225.56 18.449 0.5625 0 108.044 1000 ! ! 97.5 3.5 5.5-11.0625 10.218 0.971 13.5 11.250 0.5-10.441 0.5-8.887 62.300 179.5 ! ! ( ni " ei ) ! 2 ( ni " ei ) 2 ! ei 0 6 34 132 179 218 183 146 69 30 3 0 0.272 2 \$ 6.074 6.774 7.5 4.509 241.5 5.104 0.116 597.5 1.795 % & ln(1 ) (2 2.5 7.5625 5.0625 1.746 178.016 227.5-5.610 22.591 2353.050 1019.5-10.895 27. (c) Acceptance gap size (secs) Gi 1 2 3 4 5 No.250 279.5-1.5 ! ! 132 179 218 183 146 69 30 3 0 100.121 66.044 0.5 7.033 0.093 542.5-4.472 76.512 6.000 47.150 207.343 2.5 4.528 0.064 14.619 107.375 359.5 11.5-5.000 0.745 5.330 11.993 ) & ln(1 ) ) & 0.993 & 1.003 3.484 25.047 135.193 0.069 1.098 2.088 5.443 23.882 0.5 8.5 9.2482 As 25.3.009 Where the lognormal distribution parameter is calculated as follows: 1 2 # & ln \$ " % 2 & ln(6.050 1000 ! ! 25.5-6.009.5-12.347 1.870 51.063 176.434 1501.102 10.5-9.883 11.5-4.354 119.000 29.804 1.5-8.5 10.248 55.5-7.5 6.512 5.5-9.794 1.582 6.948 5.227 2.622 26.5-2.5 6.

0625 22.25 sec Var(G) = 2.300 0.5625 33.281 1.168 0.0625 0.5625 3.25 0.875 67.625 102.6 7 8 9 10 11 12 ! "G = 6.502x10-3 .125 521.0625 13.0625 7.5625 14.308 1.688 0 2990.729 218 183 146 69 30 3 0 1000 1.938 447.033 0 6.813 421.99 # G = 1.621 0.

52 22 54.10 -1.78 0.80 0.28 28 67.93 1.97 6 43.13 -1.87 1.40 -0.63 0.50 0.17 -0.70 0.47 -0.11 5 43.34 20 53.73 0.11 Rearrange the given data in increasing order.02 0.31 0.34 12 47.83 2 40.21 0.26 0.84 7 43.08 17 50.30 0.60 0.27 -0.28 4 42.07 -1.38 0.53 0.83 (a) Plot the data points in a lognormal probability paper .30 -0.17 14 48.7.62 23 55.93 0.57 0.50 29 67.93 0. we obtain the following table: Observed Rainfall m m/N+1 S Intensity X.05 0.12 0.37 -0.77 0.96 0.03 -1.11 0.67 0.62 9 45.20 -0.83 0.43 11 46.52 0.91 0.00 16 50.52 10 45.11 27 63.73 24 58.97 1.77 0.71 0.97 26 59.77 0.83 0.33 -0.08 15 49. in 1 39.90 1.29 0.51 0.25 19 53.49 0.91 0.37 0.43 21 54.50 3 41.57 0.72 0.67 0.28 0.17 18 51.73 8 44.84 25 58.59 0.23 -0.43 -0.25 13 48.

784 7. S (b) From the data in probability paper xm = 50.116 4.0 1.027 0.495 3.00 3.84/xm) = 0.000 0.92 The degree of freedom for the lognormal distribution is f=6-1-2=3.137 0.454 4.000 0.0 Observed Rainfall Intensity.16 ! " = ln(xm) = 3.00 1.100.852 . On the basis of the observation data &ˆ ' 50.021 0.915 % 29 29 (ni \$ ei ) 2 ei 2 0.913 0.00 2. in y = 50.16e0.947 ( ni \$ ei ) 2.92 x0.590 1.00 0.445 Perform Chi-square test for Gamma distribution Observed Theoretical Interval (in) frequency ni frequency ei <40 40-45 1 7 2.231 1.629 4.8 ! # = ln(x0.00 -1.215 2 ( ni \$ ei ) 2 ei 0.354 (ˆ 2 ' 61.118 4.054 0.179 .018 6.159 Perform Chi-square test for log-normal distribution Interval (in) Observed frequency ni Theoretical frequency ei (ni-ei) <40 40-45 45-50 50-55 55-60 >60 1 7 7 7 4 3 2.00 -2.159x 10.245 0.823 kˆ ' 41. thus )ˆ ' 0.250 4.00 Standard Normal Variate.84 = 58.862 0.0 -3.956 1.

422 0.81 .004 0. .030 7 6. Comparing with the 0.790 ! ( ni \$ ei ) 2 calculated. the Gamma distribution is superior to the lognormal distribution in this problem.95.45-50 50-55 55-60 >60 7 7. c. For a significance level " # 5% . ei both the Gamma distribution and Lognormal appear to be valid model for the rainfall intensity at the significance level of " # 5% .069 29 29 ! The degree of freedom for the Gamma distribution is f=6-1-2=3.021 1.040 0.741 0.010 0.067 4 4.178 3 3.263 0. As the ( ni \$ ei ) 2 ! e in Gamma distribution is less than i that of the lognormal distribution.173 0.3 # 7.

25> c. xm = 0.0-1.0 1.75 0.25-1.0 y = 0.2=5.0 Ratio of Settlement 3.171 0.00 0. Ratio of settlement Observed frequency ni Theoretical frequency ei (ni-ei)2 <0. Comparing with the ' normal distribution is not a valid model for the ratio of settlement.0 0.00 2.95.95.312 37.475 5. (ni \$ ei ) 2 calculated. For a significance level &=5%.5 % 2 13 8 1 1 25 5.00 3. the ei .356 (ni \$ ei ) 2 ei 2.3322x + 0.783 6. S Linear trend is observed.897 6.99 The degree of freedom for the normal distribution is f=5-1-2=2.223 10.2=5.420 8.99.25 1.246 1. c.994 ! # = slope = 0.596425 25 14.00 -1.994 1.00 -2. but there are two data points out of the trend.12 (a) Plot data on normal probability paper 4.00 Standard Normal Variate.400 0.0 2.00 1.3322 (b) Perform Chi-square test for normal distribution.499 0. (b) Based on the trend line.994 ! "= 0.0 -3.7.808 3.915 1.5 >1.209 2.75-1.

354 0. 0 A* & A2 .99 1.386 -1.86 0.614 0.14 1.507 0.960 -1.435 0.519 0.22 i &1 For normal distribution.536 -1.495 0.38 2.555 0.300 0.343 0.343 0.555 0. b1 =-0.560 -0.75 2.118 -1.253 -0.12 0.86 0.435 0.614 0.712 0.495 0.579 0.18 1.94 0.726 for a prescribed significance level 2 = 5%.6).366 0.077 0.412 0. .670 0.077 0.94 0.471 0.02 1.519 0.877 1.435 0.412 0.712 0.579 0.000 0.366 0.84 0.519 -1.000 1.321 0.300 0.628 -0.09 1.509 -1.343 0.435 0.471 0.97 0.1.04 1.495 0. .004 -0.851 -0.188 -1. i xi F(xi) F(xn+1-i) (2i \$ 1) !ln FX ( xi ) # ln[1 \$ FX ( xn#1\$i )]" / n 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 0.89 (Table A.Perform Anderson-Darling test for normal distribution. the normal distribution is not acceptable at the significance level 2 = 5%.88 0.82 0.0 # 0.089 -1.462 -1.22 Calculate the Anderson-Darling (A-D) statistic n A2 & \$+ ')(2i \$ 1) !ln FX ( xi ) # ln[1 \$ FX ( xn #1\$i )]" / n (* \$ n = 27.877 0.92 0.87 0.670 0.52 0.490 -0.531 0. the adjusted A-D statistic for a sample size n=25 is. b0 =-0. As A* is greater than c2 .008 %=-27.447 -1.495 0.531 0.7514. b b c2 & a2 .795.06 1.451 -1. .22-25 = 2.321 -1.507 0.37 0.522 -1.25 # 2 / =2.01 1.396 -0.657 -0. 1 # 0 # 12 / n n 1 0 = 0.490 -1.294 n n 1 The critical value c2 is given by.00 1.04 1.727 -0.004 0.356 -1.354 0.555 0.321 0. a2 =0.293 -1.343 0.99 0.555 0.

0 1.9143 ! " = ln(xm) = -0.734 0.783 -0.830 4.75-1. i xi F(xi) F(xn+1-i) (2i \$ 1) (ln FX ( xi ) * ln[1 \$ FX ( xn*1\$i )]) / n 1 2 3 4 0.(c) Plot the data on the lognormal probability paper.00 -1.2=5.84 0.073 10. S Therefore.451 -0.75 0. (ni \$ ei ) 2 calculated.00 3.135 38.814 25 34.95.95.843 0.096 1.4081 Perform Chi-square test for lognormal distribution.5 >1.161 1.083 0.99 The degree of freedom for the normal distribution is f=5-1-2=2. Comparing with the ' ei lognormal distribution is not a valid model for the ratio of settlement.00 -2.990 0.000 0. xm = 0.18> c. Ratio of settlement Observed frequency ni Theoretical frequency ei (ni-ei)2 <0.730 2.00 Standard Normal Variate.169 %= 13.99.992 3. Ratio of Settlement 10.521 -0.82 0.843 6.418 0.784 2.395 0.00 0. Perform Anderson-Darling test for lognormal distribution.25-1.12 0.0896 # = slope = 0.290 (ni \$ ei ) 2 ei 4.00 2.0-1.2=5.52 0.9143e0.4081x 0.706 -0.1 -3.0 y = 0.341 2.00 1. the For a significance level &=5%.0 1.5 % 2 13 8 1 1 25 7.587 .353 5. c.575 2.25 1.

527 0.00 1.624 0.577 0.593 -1.440 0.452 0.440 0.14 1.0 \$ 0.570 -1.25 \$ 2 / =3.04 1.75 2.530 -1.019 !=-28. As A* is greater than c2 .706 0.18 1.577 0.06 1.395 0.16 n n 1 The critical value c2 is given by. b0 =-0.596 0.624 0.92 0. 0 A* & A2 . 1 \$ 0 \$ 12 / n n 1 0 = 0.37 0.04 1.530 -1.440 0. the lognormal distribution is not acceptable at the significance level 2 = 5%.795.049 -1.99 1.726 for a prescribed significance level 2 = 5%.667 0.09 1.463 0. b1 =-0.617 -1.99 0.7514.990 0.281 -1.01 1.452 0.440 0.06 Calculate the Anderson-Darling (A-D) statistic n A2 & %+ ')(2i % 1) "ln FX ( xi ) \$ ln[1 % FX ( xn \$1%i )]# / n (* % n = 28.94 0.558 0.461 -0. .000 -0.577 0.487 -1.691 -0.641 0. .506 0.096 -1.94 0.410 -1. .86 0.88 0.606 0.02 1.734 0. a2 =0.624 0.87 0. b b c2 & a2 . the adjusted A-D statistic for a sample size n=25 is.083 0.176 -1.86 0.598 -1.624 0.812 -0.641 0.606 0.596 0.527 0.97 0.506 0.558 0.38 2.527 0.06-25 = 3.330 -1.587 0.922 -1.5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 0.474 -1.577 0.587 0.667 0.843 0.463 0.527 0.06 i &1 For lognormal distribution (lognormal distribution should be the same as normal).484 -0.418 0.1.587 -1.89 (Table A.6).

0 30.50 2.972=0.348 -0.584 3.993 2.067 0.593 4.019 0.2=5.0 10.728 1.2884 80. c. Mean depth (m) Observed frequency ni Theoretical frequency ei <5 5-15 15-25 25-35 35-45 % 3 4 1 4 0 15 2.072 2.99 The degree of freedom for the shifted exponential distribution is f=5-1-2=2.50 4. (ni \$ ei ) 2 calculated.95.90 4.03575 Perform Chi-square test for shifted exponential distribution.2=5.277 %= 4.0 Observed Rainfall Intensity. Perform Anderson-Darling test for exponential distribution.834 -0.950 0.173 0.95 < c.0 70.95.0 50.0 20.2884.163 0.99.0.50 3.70 0.00 4.00 2.722 .13 Plot data on (shifted) exponential probability paper.0 y = 27.50 1. in 90.00 3.0 40.630 0. the shifted For a significance level &=5%.0 60.0 0.732 2. Comparing with the ' ei exponential distribution is an appropriate model for the mean depth of Glacier Lakes.826 1.7.108 0.590 1. 100.277 15 (ni-ei) (ni \$ ei ) 2 ei 2 0.610 1.972x .00 1.00 Standard Normal Variate. i xi F(xi) F(xn+1-i) (2i \$ 1) (ln FX ( xi ) * ln[1 \$ FX ( xn*1\$i )]) / n 1 2 2. # = 1/27. S !" a = -0.0 0.00 0.50 5.

c.644 0.356 0.90 50. = 5%. .0 \$ 0.149 -1. .710 0.710 0.26-15 = 0.232 0.391 0.60 27.320 !=-15.60 33.950 0.6b.647 -0.318 0.232 0. = 1.114 -1.409 -1.60 18.26 i &1 For exponential distribution. both Chi-square test and Anderson-Darling test have shown the shifted exponential distribution is an appropriate model for the mean depth of Glacier Lakes.172 0.356 0.834 0.172 0.40 12. A* & A2 (1.391 0.699 0.387 -1.491 0.931 -0. is given in Table A.656 -0. the adjusted A-D statistic for a sample size n=15 is.10 10.491 0.699 0.00 7.259 -1. A* < c.30 0.6 ) = 0.036 -0.318 0.108 -1.3 4 5 6 7 8 9 10 11 12 13 14 15 5.815 0.635 0.077 -1.635 0.644 0.516 -1.686 -1.00 83.00 13. Therefore.30 34.30 46.3 n The critical value c. therefore the exponential distribution is acceptable at the significance level .90 28.321 for a prescribed significance level .163 0.815 0. = 5%.26 Calculate the Anderson-Darling (A-D) statistic n A2 & %+ ')(2i % 1) "ln FX ( xi ) \$ ln[1 % FX ( xn \$1%i )]# / n (* % n = 15.

5 – 3.00 40.56 5.75 30.99 2.4 1.509 x2 distribution with f = 9-3 = 6 degree of freedom and " = 1% significance level.997 0.5 – 3.06 8.07 208.812 & ei i \$1 9 In this case So normal distribution is not a suitable one.5 > 9.35 222.5 6.6 = 16.06 1000 (ni-ei)2 (ni-ei)2 / ei 81.5 – 3.135 0.812 (n i % ei ) 2 \$ 20.33 177.99.56 653. .Acceptance gap size (secs) Gi < 2.5 8.357 0.075 3.5 7.5 3.72 16.64 5.75 1002.91 100.5 – 3.5 ! Observed Frequency ni Theoritical Frequency ei 6 34 132 179 218 183 146 69 33 1000 15.076 0.57 138.5 – 3.5 5.00 47.5 – 3.167 9.015 0.5 2.82 49. c0.96 66.287 20.509 # 16.5 4.5 – 3.

1 2 3 4 5 6 # a\$ ^ "\$ bi 10 20 30 40 50 60 210 ai 9x10-4 20x10-4 28x10-4 41x10-4 52x10-4 63x10-4 213x10-4 213 % 10 &4 \$ 35.1.8. 6 )( \$ ' 2(bi & " a i )(&a i ) \$ 0 )" i \$1 or. A = strain = x / 10 ^ ^ Let.5 %10 i &3 i &8 2 2 ai2 81x10-8 400x10-8 784x10-8 1681x10-8 2704x10-8 3969 x10-8 9619x10-8 2 &4 &8 i = 9. 6 b\$ aibi 9x10-3 40x10-3 84x10-3 164x10-3 260x10-3 378x10-3 935x10-3 210 \$ 35 6 ' a b & nab \$ 935 %10 & 6 % 35 % 35.1 (a) B = stress = y / 0.5 % 10 & 4 .7 \$ 2.5 %10 ' a & na 9619 %10 & 6 % 35.5 %10 & 4 \$ 35 & 32.21 x 103 k/in2 (b) Assume E(B ^ A = a) = " a 6 (2 \$ ' (bi & "a i ) 2 i \$1 Now.21%10 3 % 35. E(B A = a) = ! + " a Nos.3 ^ So Young’s modulus = " = 9.21 x 103 k/in2 ^ ^ ! = b & " a \$ 35 & 9. ^ "\$ 'a b 'a i i 2 i \$ 935 % 10 &3 \$ 9.72 % 10 3 ksi &8 9619 %10 bi2 100 400 900 1600 2500 3600 9100 .

8..%)*%? 32 73 72 63 62 53 52 43 42 3 2 2 52 72 82 92 422 452 472 :(//./. % ! % Speed Stopping Distance (kph) (m) Xi Yi % % % % % % % % % % Xi2 Yi2 XiYi Yi'=a+bXi (Yi-Yi')2 4% 72 43 4822 553 822 46@76 5@784 5% A 5 94 7 49 4@78 2@599 6% 422 72 42222 4822 7222 68@3A 44@3A9 7% 32 43 5322 553 B32 4B@5A 3@535 3% 43 7 553 48 82 6@B9 2@279 8% 83 53 7553 853 4853 56@29 6@8B8 B% 53 3 853 53 453 B@87 8@AB5 9% 82 53 6822 853 4322 54@43 47@927 A% A3 62 A253 A22 5932 67@88 54@B33 42% 83 57 7553 3B8 4382 56@29 2@975 44% 62 9 A22 87 572 A@3B 5@78B .)'\$-*.%)*%<(= (b) Vehicle No.2 (a) !"#\$%#&%'\$#(()*+%./%0'%'(//.%#&%\$1-0/" :\$#(()*+%>)'\$-*.

!"#

#

!"\$

#

%\$

#

Total:

!\$&"\$

#

&)*

"'"\$

#

"+,

\$"&+!

\$&"\$

#
&*!'

%&("\$

#
!,*\$+

!(\$\$"

#

#

)!()!&

On the basis of calculations in the above table we obtain the respective sample means of X
and Y as,

X = 679/12 = 56.6 kph,

Y = 238/12 = 19.8 m

and corresponding sample variances,

1
(52631 # 12 " 56.6 2 ) ! 1289.84
11
1
S y2 ! (6910 # 12 " 19.8 2 ) ! 200.50
11

S x2 !

,

From Eq. 8.4 & 8.3, we also obtain,

!

\$!

18953 # 12 " 56.6 " 19.8
! 0.388 ,
52631 # 12 " 56.6 2

!

% = 19.8 – 0.388"56.6 = -2.161

From Eq. 8.6a, the conditional variance is,

S Y2| X !

1 n
( y i # y i' ) 2 = 71.716 / (12 – 2) = 7.172
&
n # 2 i !1

and the corresponding conditional standard deviation is S Y | x = 2.678 m
From Eq. 8.9, the correlation coefficient is,
n

1
(ˆ !
n #1

&x y
i !1

i

i

# nx y

sx sy

!

1 18953 # 12 " 56.6 " 19.8
! 0.98
11 1289.84 ' 200.50

(c) To determine the 90% confidence interval, let us use the following selected values of Xi = 9,
30, 60 and 125, and with t0.95,10 = 1.812 from Table A.3, we obtain,
At Xi = 9;

( )Y

X

' 0.90 # 1.33 & 1.812 \$ 2.679

(9 " 56.6) 2
1
%
# ("1.063 ! 3.723)m
12 (52631 " 12 \$ 56.6 2 )

At Xi = 30;

( )Y

X

' 0.90 # 9.48 & 1.812 \$ 2.679

(30 " 56.6) 2
1
%
# (7.708 ! 11.252)m
12 (52631 " 12 \$ 56.6 2 )

At Xi = 60;

( )Y

X

(60 " 56.6) 2
1
' 0.90 # 21.12 & 1.812 \$ 2.679
%
# (19.712 ! 22.528)m
12 (52631 " 12 \$ 56.6 2 )

At Xi = 125;

( )Y

X

' 0.90 # 46.34 & 1.812 \$ 2.679

(125 " 56.6) 2
1
%
# (43.220 ! 49.460)m
12 (52631 " 12 \$ 56.6 2 )

8.3
(a)
Plot of peak hour traffic vs daily traffic
1.6

1.4

Peak hour traffic
(1000 vehicles)

1.2

1

0.8

0.6

0.4

0.2

0
0

(b)

1

2

3

4

5

Daily traffic (1000 vehicles)

6

7

Let X be the daily traffic volume and Y the peak hour traffic volume, both in thousand
vehicles. By Eq. 8.9, the correlation coefficient between X and Y is estimated by
n

1
\$ˆ "
n !1

#x y
i

i "1

i

! nx y

sx s y

n

With n = 5,

#x y
i "1

i

i

= 25.54, x = 1.02, y = 4.72, and the sample standard deviations sx

= 0.303315018 while sy = 1.251798706,
% \$ˆ = (1/4)(1.468 / 0.379689347) & 0.967
(c)
xi
5
4.5
6.5
4.6

yi
1.2
1
1.4
0.9

xiyi
6
4.5
9.1
4.14

xi2

y i' " (ˆ ) 'ˆxi

(yi – yi’)2

25
20.25
42.25
21.16

1.085577537
0.968474793
1.436885769
0.991895341

0.0130925
0.00099384
0.00136056
0.00844475

Sum
Mean

3

0.6

1.8

9

23.6
4.72

5.1
1.02

25.54

117.66

0.61716656

0.00029469
0.02418634

Since we know the total (daily) traffic count (X), and want a probability estimate about the
peak hour traffic (Y), we need the regression of Y on X, i.e. the best estimates of ! and "
in the model
E(Y | X = x) = ! + "x
From Eq. 8.4 & 8.3
n

!

"\$

%x y
i

i \$1
n

%x
i \$1

2
i

i

# nx y
# nx

2

= (25.54 - 5&4.72&1.02) / (117.66 – 5&4.722) = 1.468 / 6.268 ' 0.234, and

!

! \$ y # "ˆ x = 1.02 – 0.23420549&4.72 ' – 0.0854
Also, by (7.5), using the results in the table above, we estimate the variance of Y (assumed
constant)

S Y2| x \$

1 n
( y i # y i' ) 2
%
n # 2 i \$1

= 0.02418634 / (5 – 2) ' 0.0081
Hence when the daily traffic is 6000 vehicles, i.e. x = 6, Y has the estimated parameters
(Y = –0.0854 + 0.234&6 = 1.319
)Y = 0.0080621140.5 = 0.089789278
Hence the estimated probability of peak hour traffic exceeding 1500 vehicles is
P(Y > 1.5) = 1 – P(Y * 1.5)
= 1 – + (1.5 # 1.319 ) = 1 – +(2.01)
0.09

'

0.0222

8.4
(a)

!"#\$%#&%'()%*+',\$+%(-()./%*#-012'\$,#-%30%'()%*+',\$+%45!
7666

!()%>+',\$+%?-()./%>#-012'\$,#-

;766
;666
:766

@

:666

@A

9766

"#B()

9666

1''()

8766
8666
766
6
6

9666

;666

<666

=666

86666

89666

!()%>+'\$,+%45!

(b)
Country # Per Capita GNP

Per Capita
Energy
Consumption

%

%

%

%

%

%

%

%

%

%

%

Xi

Yi

Xi2

Yi2

XiYi

Yi'=a+bXi

(Yi-Yi')2

8%

<66%

8666%

:<6666%

8666666%

<66666%

C76D:E%

9;<:D9<C%

9%

9E66%

E66%

E9C6666%

;C6666%

8=C6666%

87:7D<;%

<C=9=<DE:;%

:%

9C66%

8;66%

=;86666%

8C<6666%

;6<6666%

87C8D:=%

:<<9;D;E=%

;%

;966%

9666%

8E<;6666%

;666666%

=;66666%

8C7:D<=%

98;7D9:=%

7%

:866%

9766%

C<86666%

<976666%

EE76666%

8<;ED88%

E9E;89DC;;%

<%

7;66%

9E66%

9C8<6666%

E9C6666%

8;7=6666%

99==D89%

8<C<;:DC6<%

E%

=<66%

9766%

E:C<6666%

<976666%

98766666%

:8ECDC<%

;<9:;8D89:%

!"

#\$%\$\$"

&\$\$\$"

#\$'\$(\$\$\$\$

#'\$\$\$\$\$\$"

&#)\$\$\$\$\$"

%'*%+,&"

##(!(%+#\$%"

"

"

"

"

"

"

"

"

Total:

%,!\$\$"

#'!\$\$"

)*)*)\$\$\$\$

&%)&\$\$\$\$"

(((!\$\$\$\$"

"

))#!!#\$+,('

On the basis of calculations in the above table we obtain the respective sample means of X
and Y as,

X = 37800/8 = 4725,

Y = 16800/8 = 2100

and corresponding sample variances,

1
(252520000 # 8 " 4725 2 ) ! 10559285.71 ,
7
1
S y2 ! (43240000 # 8 " 2100 2 ) ! 1137142.86
7

S x2 !

From Eq. 8.4 & 8.3, we also obtain,

!

\$!

(c)

99980000 # 8 " 4725 " 2100
!
! 0.279 , % = 2100 – 0.279"4725 = 781.725
2
252520000 # 8 " 4725

From Eq. 8.9, the correlation coefficient is,
n

1
(ˆ !
n #1
(d)

'x y
i !1

i

i

# nx y

sx sy

!

1 99980000 # 8 " 4725 " 2100
! 0.85
7 10559285.71 & 1137142.86

From Eq. 8.6a, the conditional variance is,

S Y2| X !

1 n
( y i # y i' ) 2 = 2218817.515 / 6 = 369801.799
'
n # 2 i !1

and the corresponding conditional standard deviation is SY | X = 608.113
(e)

To determine the 95% confidence interval, let us use the following selected values of Xi =
600, 5400 and 10300, and with t0.975,6 = 2.447 from Table A.3, we obtain,
At Xi = 600;

588 . * " 2.113 (600 \$ 4725) 2 1 % " (62.253) 8 (252520000 \$ 8 # 4725 2 ) At Xi = 10300.37 & 2. + = -709.( )Y X ' 0.95 " 950.673 and S X |Y = 1853. ( )Y (f) X ' 0.74 & 2.263 ! 1835.407 ! 2827.447 # 608.113 % " (1749.95 " 3653.447 # 608.391 ! 4754.12 & 2.447 # 608.469) 8 (252520000 \$ 8 # 4725 2 ) ! Similarly.95 " 2288.080 for predicting the per capita GNP on the basis of the per capita energy consumption. .113 ! (10300 \$ 4725) 2 1 % " (2556. ( )Y X (5400 \$ 4725) 2 1 ' 0.997) 8 (252520000 \$ 8 # 4725 2 ) At Xi = 5400.

198442 1.33.35. we estimate the variance of Y (which is assumed constant) S Y2| x * 1 n + ( yi ! yi' ) 2 n ! 2 i *1 = 1.125) = 1 – %(1. sum average xi (kips) 2.8.0 yi (mpg) 25 17 20 21 13.33 ) P(X > 4.33641 16.347158218.125) = 1 .5) = P( " # 1.9 1.3.9 268.75 n ! (* +x y i i *1 n +x i *1 2 i i xiyi xi2 y i' * 'ˆ ) (ˆxi (yi – yi’)2 6.6 3. 8.993856 ! nx y ! nx 2 = (45.347158218).0.998 .468 / 6.0 20.00 62.4.55453 22.325 & 35.3 = 25.16283 0.4 72 63 24.993856 / (4 – 2) & 0.64 12.20 Also.3 kips.002891 0.3.e.75 – 4.3. by Eq.5 71. 1.96 9. and ! ' * y ! (ˆ x = 20. Y has the estimated parameters \$Y = 35.9969278030.997 Hence when the car weighs 2.3.5 (a) Let X be the car weight in kips.352165 45. x = 2. using the results in the preceding table. Hence X ! \$ 4.325 83.25 17.268 & .3252) = 1.130 (b) Let Y be the gasoline mileage.20430108 + (-4.5 ! 3.325.6a.2.4.9 – 4. X ~ N(3.8697 & 0. Based on the following data.94624 19.20.440358 0.04 = P(Z > 1. i.75) / (268.5 = 0. we assume E(Y | X = x) = ' + (x and seek the best (in a least squares sense) estimates of ' and ( in the model.3 3.5 4.2 3.206 #Y = 0.04). For linear regression.9 .

9 31 2 1 1 3 4 5 Weight (kip) 6 2 .998 (4.34 ) 4.303 from Table A.85 " 4 ' 3.2 = 4.5 " 3. + .*'0\$ 51 Gasoline Mileage (mpg) 42 7 41 78 "#/.8) = 1 – 0.%.('.998 At Xi = 30.9 32 :.)%/.206 ) 0.3) 2 1 ( & (13.613 % 20.5.3) 2 1 ( & (21. and with t0.Y X * 0. we obtain. let us use the following selected values of Xi = 2.Hence the estimated probability of gas mileage being more than 28 mpg is P(Y > 28) = 1 – P(Y ! 28) = 1 – # ( 28 " 25.95 & 16.5 and 4.998 = 1 – #(2.%-*". At Xi = 2.3 2 ) * 0.85 " 4 ' 3.303 ' 0.95 & 24.3 2 ) !"#\$%#&%'()#"*+.215 % 27.975.303 ' 0.465)mpg 4 (45.Y X (2.0026 To determine the 95% confidence interval...287)mpg 4 (45.9974 \$ 0.2.2 " 3.95 ) 4.3. (c) + .

so our regression model is only valid for predictions using Y values within this range.0.713 (b) No.073026652 = '(0. Hence the mean measurement error when a surveyor has 15 years of experience is estimated to be 1. 8.6 yi 3 5 10 20 25 63 yimi 4.847973713. and M be the measurement error in inches.5 0.5 47 yi2 9 25 100 400 625 1159 From Eq.5).030010953.5 4 10 16 12.030010953)(15) = 0.3.8 0. We have the following data: i 1 2 3 4 5 ! mi 1.6 (a) Let Y be the number of years of experience. such as effects of aging—a 100-year-old surveyor might not be able to see the crosshair anymore! .298138007 so our regression model is Mmean = 1.298138007 + (-0. other factors could come into play.847973713 ) 0.8.562571845) ( 0. while the estimated variance of M is 0.5 4. we have "! # % y m \$ nym = -0. In fact. hence P(M < 1 | Y = 15) = '( 1 \$ 0.073026652 (by equation 7. 60 is outside the range. where our model may not be correct.298138007 . our data for Y ranges from 3 to 25.4 & 8.8 1 0. and % y \$ ny i i 2 i 2 &! # m \$ "!y = 1.030010953Y.

"! .! #\$#&'*! #\$))&! #\$"*! #\$###)'! .#\$%'! #\$+.4 2 ^ ! % DO & " t % 0.2.3647.3092 – 0.! %! &\$.61.3092 So the least-squares regression equation is. # ti2 = 60.180. # DOi2 = 0.! +\$&! #\$"'! "#\$&)! #\$#+&)! #\$%(.4 ' 0.7 (a) Let E(DO T=t) = ! + "t T DO ! ! ! ! ! (days) (ppm) ! ! ! ! ! ! Xi Yi Xi2 Yi2 XiYi Yi'=a+bXi (Yi-Yi')2 "! #\$%! #\$&'! #\$&%! #\$#(')! #\$")! #\$&*! #\$###"&! &! "! #\$&*! "! #\$#')"! #\$&*! #\$&(! #\$###%&! +! "\$.3647 & 8 ' 0.20 ) 0.0455 60.! #\$&*! &\$%.4 2 ) % 2.! #\$"(! .)! #\$&)! #\$##&.825 x 10-4 t% ^ 19.! #\$###&(! (! +\$'! #\$"! ")\$))! #\$#"! #\$+'! #\$")! #\$##"&)! '! )\$(! #\$"&! &&\$#*! #\$#"))! #\$%. # DOi = 1.58.0058 7 .#! )! "\$'! #\$"'! +\$&)! #\$#+&)! #\$+&)! #\$&+! #\$##&).2 2 ) % 0.4.18 & 8 ' 2.! #\$#')"! #\$). ^ E (DO T=t) = 0. 8 "% DO % #t i ( DOi & nt DO ^ 2 #t i & nt 2 % 1.2 % 2.0455 t (b) S x2 % 1 (60. \$2 = #(DOi – DOi’)2 = 84.4 % 0.! #\$".)! #\$#*! #\$###()! Total: "*\$&! "\$.58 & 8 ' 2.\$(.61 % 0.)(! +\$"'! ! #\$##')&! # ti = 19.58 & 8 ' 2.20 8 3.8. 7 S y2 % 1 (0.20 % &0.0714 . # ti DOi = 3.0455 ' 2.

0058 1 n ( y i ! y i' ) 2 = 0.4 \$ 0.037 ppm.2 " !0.n 1 &ˆ " n !1 S Y2| X " %x y i "1 i i ! nx y sx sy " 1 3.00842 / 6 = 0.89 7 2. .18 ! 8 \$ 2.0714 # 0.0014 % n ! 2 i "1 and the corresponding conditional standard deviation is SY | X = 0.

4 + 0.9995 (c) Let YA and YB be the respective actual strengths at A and B. Since these are both normal.999963 (b) Theorem of total probability gives P(Y > 0.12!0.3) = P(Y > 0. YB ~ N(0.4) = 0.3 | X = 0.4) 0. When X = 0.442 – 0. hence P(Y > 0. 0.4) = 1 – P(Y " 0.05 2 ).442.498. Hence the difference D = YA – YB ~ N(0.9977!(1/5) + 0.05. if X = 0.442 ) = 0. the mean of Y is E(Y | X = 0.8.05 2 ' 0.3 | X = 0.35) = 1 – P(Y " 0.498 % P(Y > 0.4 % E(Y | X = 0.35.05 Also.4) = 1.3 | X = 0.9977 = 1 – #( 0.442. i.005 ) Hence P(YA > YB) = P(YA – YB > 0) = P(D > 0) 0. 0.05 = 0.056.12!0. D ~ N(–0.999963!(4/5) & 0.05). YA ~ N(0.3 \$ 0.8 Cost vs floor area 250 thousand \$ 200 150 100 50 0 0 (a) 1 2 3 1000 sq ft The standard deviation of Y is 4 0.35 + 0. as preparation for part (b).35)P(X = 0.498 ) = 1 – #( 0.0025 = 0.498.3 | X = 0.35) 0.35) + P(Y > 0.05 = 0.e.05).4)P(X = 0.35) = 1.3 | X = 0.3 \$ 0.05 = 0. .3 | X = 0. 0.

785807948 # 0.= 1 – ![ 0 " ("0.056) ] 0.791959595) = 1 – 0.214 .005 = 1 – !(0.

sy = 74. we should not simply calculate the sample standard deviation of the y data.5 & 19.311 / - 2.05 63 1. assuming it is a constant.46028024.246093 33.36 640.335 -19.23086667 139.10a. (b) Note the words “for given floor area”.00666667 119.666667 xiyi xi2 66.02 359 Mean = 2. since although the individual yi’s may deviate a lot from their mean.1928250. . In this case.465 The regression line y = –19. Plugging these (and beta) into 8.33478768 x can be plotted on the scattergram.1074492 364.8.8596 14.699949 198. and has the best fit to the data (in a least square sense).07 1. this may be due to changes in x and hence not really a randomness of y.056140773.3854267 107.3363865 93.14 204 Sum = 6.15 168.08 (c) The individual sample standard deviations are sx = 1.131807 yi' (yi-yi')2 53.676667 119. Hence we need to calculate the conditional variance E(Y | x).192825 / (3 – 2) & 364.3489 9.1025 3.46514060 + 69.9 (a) The formulas to use are n #x y ! \$" i i "1 n #x i "1 2 i ! nx y i ! nx 2 and ! % " y ! \$ˆ x The various quantities involved are calculated in the following table: n= 3 xi yi 1.417521 237.666667 Beta = Alpha = 154. 8. An unbiased estimate of this is given by Eq.56 875.6a as S Y2| x " 1 n # ( yi ! yi' ) 2 n ! 2 i "1 = 364.192825 = = 69.193 Thus the standard deviation of construction cost for given floor area is estimated as SY|x = 364.83 92 3.

8718286.46028024 & 0.33478768(25) & 153.983 Since it is close to 1. thus Y ~ N(153.3347877 \$1.0838367 5 .915 19.5.056140773%74. 19.8718286."ˆ # !ˆ Sx Sy = 69. 'Y = ( + !(25) = –19. there is a strong linear relationship between X and Y (d) When X = 2.5) = P(Z < 180 * 153.46514060 + 69.871828 6 ) = +(1.369125704) & 0.08383675) ) P(Y < 180 | X = 2.

3.4 & 8. X =145/6 = 24.7 ! 0.6a.17 & 20.! %#'! (#&! &'\$! "*)#*! ")%+"! %! "'! "#! ##'! "&&! ".7 and corresponding sample variances. the correlation coefficient is.9.512 2 3675 # 6 " 24. % = 18. S x2 ! 1 (3675 # 6 " 24.2 2 ) ! 34.10 (a) Car # Rated Actual Mileage Mileage ! ! ! ! ! ! ! ! (mpg) (mpg) ! Xi Yi Xi2 Yi2 XiYi Yi'=a+bXi (Yi-Yi')2 "! #\$! "%! &\$\$! #'%! (#\$! "')'&! \$)#"'! #! #'! "*! %#'! (%"! &+'! "*)#*! \$)\$.8.751 .\$! "")+.2 " 18. 8. the conditional variance is.\$+! &! (\$! ##! *\$\$! &.2 From Eq.751"24. 5 1 S y2 ! (2194 # 6 " 18. n 1 (ˆ ! n #1 'x y i !1 i i # nx y sx sy ! 1 2835 # 6 " 24. ! \$! (b) 2835 # 6 " 24. 8.7 ! ! 0.67 From Eq.2.! \$)\$&.67 5 From Eq. .! ! ! ! ! ! ! ! ! Total: "&'! ""#! (%+'! #"*&! #. we also obtain. 8.7 2 ) ! 20.17 .7 – 0.('! ! %)*#+! On the basis of calculations in the above table we obtain the respective sample means of X and Y as. Y = 112/6 = 18.97 5 34.2 = 0.&! %%\$! #()\$'! ")"\$\$! '! #'! ".2 "18.%! (! (\$! #'! *\$\$! %#'! +'\$! #()\$'! ().

P(YQ > YR) = P(YR .512 + 0. (c) YQ = actual milage of model Q car YR = actual milage of model R car ! XQ = rated milage of model Q car = 22 mpg E(YQ) = 0.03. YR = (18. YR to be statistically independent. 1.732 n # 2 i "1 and the corresponding conditional standard deviation is SY | X = 1.54 # 17.32 2 " \$ (#0.316 mpg.03 Similarly E(YR) = 0.809) " 0.03) 1.927 / 4 = 1.751x24 = 18.54 Hence YQ = N(17.32).512 + 0.32) Assume YQ. 1.YQ < 0) = \$ # (18.21 .751x22 = 17.54.S Y2| X " 1 n ! ( y i # y i' ) 2 = 6.

132 from Table A.78 & 2. ( )Y X (20 \$ 24. we obtain.90 " 23.132 #1.316 (30 \$ 24.2) 2 1 ' 0.331 ! 24.114) mpg 6 (3675 \$ 6 # 24. 20 and 30. let us use the following selected values of Xi = 15.446 ! 14.4 = 2.90 " 11.2 2 ) At Xi = 30.05 & 2.2) 2 1 % " (21.2 2 ) .014) mpg 6 (3675 \$ 6 # 24.316 1 (15 \$ 24.3.2) 2 % " (9. ( )Y X ' 0.95.769) mpg 6 (3675 \$ 6 # 24. ( )Y X ' 0.54 & 2.90 " 15.132 #1. At Xi = 15.066 ! 17.(d) To determine the 90% confidence interval.316 % " (14. and with t0.132 #1.2 2 ) At Xi =20.

.11 (a) Case No."\$#)! )\$+)'! ")! )(\$)! *"\$-! '-)\$+&! (()\$)-! ("(\$'! )&\$)+! -)\$&'&! "*! *)! *"! "#)&! (+"! (()! )+\$%*! "'\$)**! "&! -\$&! *\$%! )(\$"+! "*\$+(! "(\$('! *\$))! #\$))(! "-! *\$+! &! ")\$(+! "+! "&\$&! "\$+*! -\$+"-! "+! *-\$(! )-\$(! ")''\$'"! +%#\$'"! ()(\$'"! *#\$"'! "'\$)("! "%! ""\$+! ""\$*! "*&\$-+! ")%\$+(! "*"\$#'! '\$%#! +\$%-%! "'! ")\$%! "&\$)! "+"\$)(! )#"\$+&! "'#\$*&! (\$+%! )#\$&(-! "(! &+! &)\$"! )""+! "%%)\$&"! "(*+\$+! *(\$"#! '\$('"! )#! *\$'! "\$-'! "&\$&&! )\$&(+&! +\$##&! "\$'"! #\$#-)! )"! &\$(! -! )&\$#"! )-! )&\$-! )\$%'! &\$(*)! ))! ""\$%! ""\$'! "*+\$'(! "*(\$)&! "*'\$#+! '\$%(! (\$#++! )*! "\$'*! *\$*(! *\$*&'(! ""\$&()"! +\$)#*%! #\$#%! ""\$#&(! )&! (\$&*! *\$)&! ''\$()&(! "#\$&(%+! *#\$--*)! +\$%'! ")\$--)! )-! +\$+! &\$*! &*\$-+! "'\$&(! )'\$*'! &\$)'! #\$###! ! ! ! ! ! ! ! ! Total: +##\$&! &("\$'! &'#+*\$"+ *'"*'\$-" &"-&+\$-" ! )"')\$(+%! On the basis of calculations in the above table we obtain the respective sample means of X and Y as. Cal.8.#\$(*! *)\$'+&! )! +&! )-\$"! &#(+! +*#\$#"! "+#+\$&! --\$#"! '(&\$+("! *! -! *\$'! )-! "&\$&&! "(! )\$'%! #\$'+(! &! )-! )+\$&! +)-! +(+\$(+! ++#! )#\$-&! *&\$)((! -! )(\$-! )%\$'! '%#\$)-! %%)\$'&! ')#\$"! )&\$-)! "#\$%--! +! *\$'! #\$-! "&\$&&! #\$)-! "\$(! "\$'"! "\$%#'! %! -\$(! +\$-! *&\$'"! &)\$)-! *'\$*-! *\$++! '\$#&'! '! *'\$"! )\$+! "&-"\$+"! +\$%+! ((\$#+! *)\$")! '%"\$&('! (! "'-! "%&! *&))-! *#)%+! *)"(#! "+"\$(-! "&-\$"(-! "#! )'\$"! )+\$%! %'(\$+"! %")\$'(! %-#\$)%! )*\$)'! ""\$+%&! ""! #\$+! #\$+! #\$*+! #\$*+! #\$*+! . S'mt. S'mt. ! ! ! ! ! ! Yi Xi Yi2 Xi2 XiYi Xi'=a+bYi (Xi-Xi')2 "! #\$%! &\$'! #\$&(! )*\$#&! *\$*+! . Obs.

551 Hence. 185. .51 # 25 "19. .. 11. E(X | y) = % + \$y = -1.54 + 2.511 ) 24.6.9.8/25 = 19. 0.95 ! 55.069 from Table A.Y = 600.9. At Yi = 64.16 # 25 " 24 2 ) ! 1401.884y From Eq.975) 25 (48063.16 # 25 " 24 2 ) At Yi = 25.069 " 9..742 (64 # 24) 2 1 * ! (49.048 ) 60.9. the conditional variance is.16 # 25 " 24 2 ! % = 19.16 # 25 " 24 2 ) At Yi = 5. X = 491. ! \$! 41546.6a.51 # 25 " 24 " 19. S X2 |Y ! 1 n ( x i # x i' ) 2 = 2182. we obtain.4 & 8.6 and 46.7 – 0.742 * ! (16.98 24 S Y2 ! . 5. n 1 (ˆ ! n #1 (c) &x y i !1 i i # nx y sx sy ! 1 41546. 1 (48063. 8.4/25 = 24.6.95 ! 20. From Eq. let us use the following selected values of Yi = 64.576) 25 (48063.7 2 ) ! 1185.912 & n # 2 i !1 and the corresponding conditional standard deviation is S X |Y = 9. X Y .551 + 0.3. . the correlation coefficient is.742 (b) From Eq.884"24 = -1.069 " 9.01 + 2. 8. 0. 3.975.967 / (25 – 2) = 94.7 and corresponding sample variances.7 " 24 ! 0.91 To determine the 95% confidence interval.7 ! 0.91 24 1 S X2 ! (38138.X Y (25 # 24) 2 1 .98 ' 1401. 48063. 25.96 24 1185. 8. 0. we also obtain.23 = 2.3. and with t0.884 .51 # 25 "19.

069 \$ 9.742 (5.70 & 2.63 & 2.159) 25 (48063. ( )X Y (0.761) 25 (48063.957) 25 (48063.069 \$ 9.069 \$ 9.983 ! 6.069 \$ 9.16 " 25 \$ 24 2 ) At Yi = 11.16 " 25 \$ 24 2 ) At Yi = 0.742 (185 " 24) 2 1 % # (143.6 " 24) 2 1 ' 0.95 # 39.742 (46 " 24) 2 1 % # (34.95 # 1.66 & 2.02 & 2.16 " 25 \$ 24 2 ) At Yi = 46.9 " 24) 2 1 % # ("0.742 (3.6. ( ) X Y ' 0.10 & 2.( ) X Y ' 0.16 " 25 \$ 24 2 ) .742 % # (4.804 ! 3.16 " 25 \$ 24 2 ) At Yi = 185.445 ! 12. ( ) X Y ' 0.95 # "1.244) 25 (48063.742 % # ("5. ( ) X Y ' 0.6 " 24) 2 1 % # ("2.6.094) 25 (48063.95 # 3.069 \$ 9.806 ! 180.833 ! 8. ( )X Y (11.403 ! 43.95 # 8.6 " 24) 2 1 ' 0.16 " 25 \$ 24 2 ) At Yi = 3.6.95 # 161.069 \$ 9.95 & 2.803) 25 (48063.

\$%/. S'mt.!"#\$%#&%'("')"(\$*+%.*1/*+%. 534 7 78 "#9*1 )::*1 544 34 4 4 34 544 234 Obs.*\$\$"*-*. S'mt.\$ 644 Cal. 534 644 .%#0.*\$\$"*-*.

7 "#<-+ 6 =//-+ 5 4 3 3 83 43 93 53 % Fare Increase (b) Fare Loss in Increase Ridership % % % % % % % % % % % % % Xi Yi Xi2 Yi2 XiYi Yi'=a+bXi (Yi-Yi')2 8% >% 8?>% 4>% 4?4>% @?>% 4?3>% 3?933% 4% 9>% 84% 844>% 855% 543% 88?>>% 3?434% 9% 43% @?>% 533% >6?4>% 8>3% 6?73% 3?5A8% 5% 8>% 6?9% 44>% 9A?6A% A5?>% >?44% 8?8@6% >% 5% 8?4% 86% 8?55% 5?7% 8?@9% 3?474% 6% 6% 8?@% 96% 4?7A% 83?4% 4?96% 3?554% @% 87% @?4% 945% >8?75% 84A?6% 6?8@% 8?3@3% 7% 49% 7% >4A% 65% 875% @?@>% 3?369% A% 97% 88?8% 8555% 849?48% 548?7% 84?>3% 8?A64% 83% 7% 9?6% 65% 84?A6% 47?7% 9?33% 3?964% 88% 84% 9?@% 855% 89?6A% 55?5% 5?4@% 3?943% .)/%0(%'%&1+-%)*2+-1(86 % Loss in Ridership 85 84 : 83 :.8.12 (a) !"#\$%#&%'%"#((%)*%+).-+(.

3.96 & 10. On the basis of calculations in the above table we obtain the respective sample means of X and Y as.3 From Eq.6a.3 = 0. n 1 (ˆ ! n #1 'x y i !1 i i # nx y sx sy ! 1 1944.6 and corresponding sample variances. ! \$! (c) 1944.3 " 5.464 2 5757 # 16 "16.6 2 ) ! 10. the correlation coefficient is. X =261/16 = 16.23 # 16 " 5.7 # 16 "16.1/16 = 5. S x2 ! 1 1 (5757 # 16 " 16.417 / 14 = 0. S y2 ! (667..&+%)# !*# !\$# )&)# "'(# !(&*%# \$)&'# +&'+# "&!.3 " 5.. the conditional variance is. 8.96 . % = 5.&!# +\$+\$&.317 .!)# !%# "*# '# +"(# %)# !')# \$&\$+# .&"+# +'&+# )&+'# . .%*# # # # # # # # # # # # # # # # # Total: "%!&.673 ' n # 2 i !1 and the corresponding conditional standard deviation is SY | X = 0.&.\$# !+# \$# "&'# )(# \$&')# !(&%# "&%'# .66 15 15 From Eq.820 (d) To determine the 90% confidence interval.66 From Eq.6 ! 0.3 2 ) ! 99.!"# !\$# %&%# "'(# )*&+%# !!"&"# +&'+# .4 & 8.317"16.6 – 0.3.&. 8. Y = 90. S Y2| X ! 1 n ( y i # y i' ) 2 = 9. we also obtain.6 ! ! 0.9.# !)# !*# )&+# !%(# ". 8. # (&)!\$# %%\$&"*# !())&\$.# (.97 15 99.&.7 # 16 " 16. let us use the following selected values of Xi = 4..

95.90 " 12.761# 0.14 = 1.311 ! 8.3 2 ) At Xi = 12.761# 0.147 ! 2.3) 2 1 % " (2.23.3 2 ) At Xi = 38.761 from Table A. ( )Y X ' 0. we obtain.68 & 1.3.27 & 1.761# 0.3 2 ) ' 0.3) 2 1 % " (11.3) 2 1 % " (3.761# 0.73 & 1.3 2 ) ' 0.82 (12 \$ 16. ( )Y X ' 0. ( )Y X At Xi = 7. At Xi = 4. and with t0.661) 16 (5757 \$ 16 #16.75 & 1. ( )Y X .387) 16 (5757 \$ 16 #16.761# 0.90 " 7.3 2 ) At Xi =23.3) 2 1 % " (1.183) 16 (5757 \$ 16 #16. ( )Y X (23 \$ 16.90 " 1.315) 16 (5757 \$ 16 #16.82 (4 \$ 16.90 " 4.82 (7 \$ 16.5 & 1.82 (38 \$ 16.90 " 2.615 ! 13. 38.181 ! 3.188) 16 (5757 \$ 16 #16.3) 2 1 ' 0. 12 and 7.870 ! 4.82 % " (7.

5 + 15x8 = 130.5 Hence P(D>150) = 1 " # (b) (150 " 100.8. E(D) = 10.5 + 15x6 = 100.65) ! 1 " 0.95 ! 0.5x0.6 + 115.3 +130.1 = 108 million \$ .1 = 100.3 + E(D\$7)x0. E(D) = 10.5x0.5x0.5 Hence E(D) = E(D\$6)x0.5 + 15x7 = 115.05 30 Given I = 7.13 (a) Given I = 6. E(D) = 10.5) ! 1 " # (1.5 Given I = 8.6 + P(D\$7)x0.

8.! )(\$&'! "*+\$#(! ! '\$%%'! On the basis of calculations in the above table we obtain the respective sample means of X and Y as.3 = -0. X =29.6a. Y = 15. 235.3 " 3.4 & 8.2/4 = 3.49 # 4 " 7.70 # 4 " 3. the conditional variance is. S Y2| X ! 1 n & ( y i # y i' ) 2 = 0. . we also obtain.220 n # 2 i !1 and the corresponding conditional standard deviation is SY | X = 0.14 (a) Load Deflection ! ! ! ! ! tons cm ! ! ! ! ! ! Xi Yi Xi2 Yi2 XiYi Yi'=a+bXi (Yi-Yi')2 "! #\$%! %\$#! &'\$()! *+\$'%! %'\$+*! %\$%%! '\$"*)! *! )\$&! *\$.8.8 2 ) ! 2.! %%\$#.96 .\$%+! +\$%+! '\$*&)! +! %\$'! *\$'! ")! %! #! "\$#"! '\$'+&! %! "'\$*! (\$(! "'%\$'%! +'\$*(! ()\$"! (\$(*! '\$''"! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! Total: *.49 # 4 " 7. 1 S x2 ! (235.591 From Eq. ! \$! 123.469 cm.3.3 2 ) ! 6.8 ! 0. 8.3.! #\$%"! ".599"7.44 / 2 = 0.8 – 0.\$+! "(\$*! *+(\$%.599 .3/4 = 7. 3 1 S y2 ! (65.3 2 ! % = 3.85 # 4 " 7.8 and corresponding sample variances.65 3 From Eq.

3 2 ) At Xi =10.2 = 2.3. ( )Y X ' 0.675 0.75) " 0.49 \$ 4 # 7. At Xi = 4. and with t0.20 cm Let y75 be the 75-percentile reflection P(Y < y75) = 0.469 and y 75 " 4.3) 2 1 % " (0. let us use the following selected values of Xi = 4 and 10.625) cm 4 (235.(b) To determine the 90% confidence interval.95.49 \$ 4 # 7. E(Y) = -0.2.90 " 5.2.20 " * \$1 (0.017) cm 4 (235.75 = * Hence y 75 \$ 4.469 % " (4.92 # 0.92 from Table A.52 & 2.90 " 1.597 ! 3.469 y 75 \$ 4. ( )Y (c) X (10.92 # 0. we obtain.81 & 2.422 ! 6.2 \$ 7.469 (4 \$ 7.52cm .3 2 ) Under a load of 8 tons.591 + 0.20 0.3) 2 1 ' 0.599x8 = 4.

"*! #")\$\$! +).! .2. 8.! *+)"\$! ()(%"! &! ''! &&! &\$&! "+%#! +#\$! &.15 (a) Deformation Brinell Hardness ! ! ! ! ! mm kg/mm2 ! ! ! ! ! ! Xi Yi Xi2 Yi2 XiYi Yi'=a+bXi (Yi-Yi')2 "! #! #\$! %#! &#'&! &(\$! #\$)#*! ()&"+! '! ""! #*! "'"! &''*! .9.8.)((! +)(((! *! '\$! %. 5 S y2 ! 1 (16659 # 6 " 50. we also obtain.97 . 1 S x2 ! (2819 # 6 " 19.4 & 8.#. X =115/6 = 19. 8. Y = 305/6 = 50.8 ! #0.99 5 122. ! 1 5014 # 6 "19.97 .8 2 ) ! 230.3.\$&! "%#+! "(%#! %\$)\$\$! %)*&%! #! %*! %'! "''*! "('&! ""'(! '+)&"! #)#++! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! Total: ""*! %(*! '\$"+! "##*+! *("&! ! '+)&"'! On the basis of calculations in the above table we obtain the respective sample means of X and Y as.2 " 50.'(! %! "%! *+! "#+! %&\$"! .2 2 ) ! 122.97 5 From Eq. n 1 &ˆ ! n #1 (b) %x y i !1 i i # nx y sx sy From Eq.97 \$ 230.8 and corresponding sample variances. the correlation coefficient is.

S Y2| X " 1 n ( y i ! y i' ) 2 = 29.412 / 4 = 7.2 # 50.71 = .712 kg/mm2. 8.353x20 = 49.544 2.7 Hence.353#19. P(40<Y ( ! 50) 50 ! 49. (c) E(Y'X=20) = 76.6a.71 2.8 + 1.2 = 76. the conditional variance is.2 From Eq.11) ! ( (!3.58) " 0. % = 50.765 2 2819 ! 6 #19.353 & n ! 2 i "1 and the corresponding conditional standard deviation is SY | X = 2.7 40 ! 49.0002 " 0.! \$" 5014 ! 6 #19.8 ! " !1.353 .7 !( " ( (0.765 – 1.544 ! 0.

5 2 ) ! 313. Y = 513/10 = 51.\$! )"(&)! %%(\$.8.&! "&''! \$.\$! #'#\$! \$&#\$! **. n 1 &ˆ ! n #1 %x y i !1 i i # nx y sx sy ! 1 20665 # 10 " 29..! "'! #'! *#! %''! "'#%! &%'! *#(&)! '(%&\$! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! Total: #+\$! \$"*! ""\$#\$! *.5 " 51.61 .9..3 and corresponding sample variances. 8.)! .99 9 313.01 9 9 From Eq.(*.3 ! 0.01 .! \$! "'! "&! "''! #\$&! "&'! "*(')! )(\$'#! &! %\$! ..61 \$ 1232."())! "&(++*! +! %\$! +'! #'#\$! )"''! %'\$'! )"(&)! &+(#. S y2 ! (37405 # 10 " 51.%'\$! #'&&\$! ! #%.3 2 ) ! 1232. the correlation coefficient is. X =295/10 = 29.&! *'%'! . S x2 ! 1 1 (11525 # 10 " 29.(*.5.16 (a) Car # Travel Speed Stopping Distance ! ! ! ! ! ! mph ft ! ! ! ! ! ! Xi Yi Xi2 Yi2 XiYi Yi'=a+bXi (Yi-Yi')2 "! #\$! %&! &#\$! #""&! ""\$'! %#(%)! "#(*)%! #! \$! &! #\$! *&! *'! *(#+! .'! *! &'! ""'! *&''! "#"''! &&''! """('.! "("\$#! %! *'! %&! +''! #""&! "*)'! \$#(#)! *+(%*.(\$*&! On the basis of calculations in the above table we obtain the respective sample means of X and Y as.! )! %'! .! "\$! "&! ##\$! #\$&! #%'! ##())! %.

5.We can say that there is a reasonable linear relationship between the stopping distance and the speed of travel. "y i ! 492 "( x i \$ x i ) 2 ! 2822. (b) !"#\$%#&%'\$#(()*+%.06 # 10 4 "( xi \$ x i )( x'i \$ x') ! 177387. So E(Y let x’ = x2 X) = a + bx + cx’ "x i ! 295./.)'\$-*./%0'%\$1-0/"%'(//. "( x' i \$ x') 2 ! 1186. 732 >\$#(()*+%?)'\$-*. "x' i ! 11525.5. "( xi \$ x i )( yi \$ y ) ! 5641 "( x'i \$ x')( yi \$ y ) ! 35829.%)*%<(= (c) E(Y X) = a + bx + cx2 .5 92 52 :2 .%)*%&\$ 722 62 52 42 32 2 2 72 32 82 42 Travel s(//..

064475 #10 7 ! 0.053x502 = 90.147 – 6.2 ! ( &1 (0.5.5 5641 177387.2 ) ' 0 . sY X ! 7.01022 # 10 9 177387.01022 #10 9 ^ " = 49.055 So E(Y X) = -6. " ! y ! 49.5 = 5641 and bx177387.28 7.5 3.9 7.0053x1152.551 ! 53.^ x ! 29.666 2.5 177387.66x50 + 0.06 # 10 4 2822.055 + 1.0053 x2 \$2 = %(yi – yi’)2 = 376.5 1186.2 Let y90 be the distance allowed P(Y < y90) = ( ( y 90 & 90.2 6x2822.6 ft .551 s Y2 X ! (d) 376.5 358295 2. the expected stopping distance is E(Y) = -6.9) ! 1.2 ! 99.334 #1.055 + 1.28 ) 90. x' ! 1152.34859 #10 9 ! ! 1.5 + cx1186.01022 #10 9 ! 1.2 – 49.5 358295 1186.793 10 & 2 & 1 .5.334 Hence.10825 = -6.334 and y 90 ! 7.5 + cx177387.5 = 49.666x + 0. y 90 & 90.06 # 10 4 2822.2 – 1.5 – 0.666x29.0053 2.334 ft At a speed of 50 mph.06x104 = 358295 det b! det det c! 5641 177387.

&#-! .3.&*'! ""+#%%%! "'&%%! "&'+#! +! "#%\$%%%! "*&+! "''%%%%%%%%! ..'#-%%&%% Yi'=a+bXi (Yi-Yi')2 On the basis of calculations in the above table we obtain the respective sample means of X and Y as.2 and corresponding sample variances.)%%%%%%%%! )%&*'! ')*%%%! *&+*! %&)"%! '! -%\$%%%! "#&*! *"%%%%%%%%! ").&*! .%%%&%! --&. 6 1 S y2 ! (2084.+\$%%%! ##&.%"%+%%! ##&"'! %&%#+! . % = 14. 1 (78469000000 # 7 " 91000 2 ) ! 3417000000 .8.72 6 S x2 ! From Eq.! %&%++! .266 .4 & 8.000181 .2.17 (a) Population Total Consumption ! ! ! ! ! 106 gal/day ! ! ! ! ! ! Xi Yi Xi2 Yi2 XiYi "! "#\$%%%! "&#! "''%%%%%%! "&''! "''%%! (%&"%! "&)*#! #! '%\$%%%! +&#! ")%%%%%%%%! #.%&#*! "&'-*! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! )&"*+! Total: ).! "*##+%%%%%%! '-. 8. X = 637000/7 = 91000.2 2 ) ! 112."&+! .#'%%%%%%%%! --#&#+! +).*')-%%%%%%&%% #%*'&-+ "#.3/7 = 14.95 # 7 "14. Y = 99.'#&#+! ###%%%%! "-&'.! .000181"91000 = 2 78469000000 # 7 " 91000 .! )%\$%%%! .%%%%! .! %&*)#! )! ". we also obtain.! "*%\$%%%! . ! \$! 12742900 # 7 " 91000 "14.2 ! ! 0.&%'! #%*%%%! '&-.2 – 0.

and with t0.6a.14 1.112 (12000 # 91000) 2 1 & " (#2. 90000 and 180000.10 ' 3.406) 7 (78469000000 # 7 % 91000 2 ) At Xi =90000.112 (90000 # 91000) 2 1 & " (12.99.98 " #0.8 ) = 1 – .185 / 5 = 1.8 P(Y > 17+X=100. ) *Y X ( 0.554 \$ 32.000) = 1 – .266 + 0. At Xi = 12000.590 \$ 15. ) *Y (d) X ( 0. ) *Y X ( 0.600 \$ 2.365 from Table A.5 = 3.(b) From Eq.365 %1.3.112 (180000 # 91000) 2 1 & " (27.98 " 30.(1.112 .08) = 0.365 %1.999) 7 (78469000000 # 7 % 91000 2 ) E(Y+X=100. we obtain (in the unit “x106 gal/day”). (17 # 15.000) = -2.98 " 14 ' 3.28 ' 3.00181x100000 = 15. the conditional variance is.420) 7 (78469000000 # 7 % 91000 2 ) At Xi = 180000.112 (c) To determine the 98% confidence interval. S Y2| X " 1 n ! ( y i # y i' ) 2 = 6. 8.237 n # 2 i "1 and the corresponding conditional standard deviation is SY | X = 1.365 %1. let us use the following selected values of Xi = 12000.

0728 \$ 0. &y i \$ %0.5 '10 % 4 \$ 5076.5 25388. ln H \$ X 2 Then Y = !0 + !1X1 + !2X2 To find !0.5 ' 10 % 4 # ! 2 25388. x 2 \$ 1.5 % 558.92 \$ 1.92 4889 % 558.H) = ln " # ! 1 ln V # ! 2 ln H Let ln E(X V.67. y \$ %0.89 25388.35 '10 % 4 ^ ^ % ! 1 ' 558.021 &( x1i % x i ) 2 \$ 4889 '10 % 4 .639. ln V \$ X 1 .9 '10 % 4 &( x1i % x i )( x 2i % x 2 ) \$ %558.075.35 ' 10 %4 &( x 2i % x 2 )( y i % y ) \$ %38680.0 '10 % 4 % ! 2 558. &x 2i \$ 19.9.2381'10 8 .92 ' 10 % 4 \$ %38680. &( x1i % x i )( y i % y ) \$ 5076. !1.866 1. ln E(X V.00175 ^ ^ ! 1 ' 4889.18 E(X V.5 % 38680.8.5 ' 10 %4 .H) = "V !1 H !2 Taking the logarithm on both sides of the above equations.89 ' 10 % 4 ^ !1 \$ 5076.89 ' 10 %4 x 1 \$ 1.H) = Y ln " \$ ! 0 . &( x 2i % x 2 ) 2 \$ 25388. and !2 &x1i \$ 12.35 % 558.

! ! #+\$.# %\$%%.618 So % = 4. '\$)&! '\$"'! #\$#(" #\$'& %\$%%." %\$%+& "" '\$*)! *\$+! %\$)++ %\$)% %\$%%# "' '\$+"! (\$'*! %\$#.89 1.35 ! 558.* %\$(+ %\$%(.! +\$((! %\$&)# %\$.50 1.2381#10 8 ! 1. "\$.53 Now ln % = B0 ^ So % = e1. * #\$))! '\$*#! '\$&*) '\$(.075 + 1. %\$%"' "% )\$(+! )\$+)! "\$+(. H " ! "+\$*%* ! ! &2 0.00175 – 0.27.866 and \$ 2 = -1.8628 #10 8 " " !1." %\$*) %\$%%" ) '\$(.! (\$")! %\$)*( %\$&" %\$%)& + '\$&. ( '\$()! &\$"&! "\$"'* %\$+( %\$#'& & '\$*'! ""\$)"! %\$'. S X V .50 ^ ^ ! ! ! ^ Mean Velocity Mean Depth Mean Oxygenation Rate (fps) (ft) (ppm/day) Vi Hi Xi ! ! ! ! Xi' (Xi-Xi')2 " #\$%&! #\$'&! '\$'&' '\$%( %\$%)) ' #\$(*! +\$%*! "\$)) "\$'+ %\$%#.866x1.26 ppm / day n ! k !1 12 ! 2 ! 1 %\$("' .639 = 1.5 ! 38680." %\$#% %\$%%% .^ \$2 " 4889 5076. 8.53 = 4. # '\$"! )\$)'! %\$*. \$ 1 = 0.612 " " 0.*! (.\$+)! ! Total: ! From Eq.2381#10 8 ^ \$ 0 = -0.5x1.618.

9 2.' .6 26.65 58.0 3.36 2.3 3.08 0.2 1. E(Y X) = 7595.31 5.5 + 6.92 4.25 1.73 -14 -11 -10 -7 40.73 5.65 4.16 1.23 1.38 0.01 7.67 &10 \$8 \$ (75.76 & 10 \$7 \$ 76.41 4.0394 0.29 1.05 4.0185 0.71 10.95 3.03 36.0609 0.6 2.41 3.06 0.8.0 0.15 0.2 0.0271 0.34 2.3523x10-6 .0 0.0010 0.47 0.03x10 157.89 3.87 6.81 8. wi (yi ^ 2 yi wi wi xi wi yi wi xi yi wi xi xi (x103) (x104) (x10-14) (x10-11) (x1010) (x10-7) (x10-8) 1.23 4.1 0. ^ %! ! Thus wi ! 1 xi 4 #wi (#wi x i y i ) \$ (#wi y i )(#wi x i ) 2 #wi ( #wi x i ) \$ ( #wi x i ) 2 40.54 7.02 14.1 3.1 0.18 &10 \$14 & 166.18 & 10 \$14 & 157.60 2.03 & 10 \$10 \$ 6.75 2.22 1.19 (a) Var (Y X ) ! " 2 x 4 .1 2.67x10-8 # So.27 9.39 9.77 3.47 2.0114 0.45 17.% xi)2 (x10-6) 0 1.0587 0.6 4.8 0.31 51.2419 0.2 0.02 2.01 2.4 3.08 4.72 2.60x10 76.4 2.9 4.08 3.01 0.65 8.35 1.22 6.7 2.2 2.6 1.41 5.60 20.72 4.6 1.8 0.44 1.2599 0.88 5.76x10 166.14 11.67 10.60) 2 & 10 \$ 22 = 6.66 2.60 & 10 \$11 40.18 & 10 \$14 ^ = 7595.03 & 10 \$10 & 75.44 4.1204 0.18x10 75.18 13.09 3.02 X ^ .3 4.60 & 10 \$11 ! '! #wi 40.44 41.02 & 75.02 ^ #wi y i \$ % #wi x i 76.4 1.5 ^ ^ To determine ' and % .35 2.64 11.82 21.77 0.4133 0.10 4.25 7.93 3.84 5.26 4.1 5.0010 0.55 2.4 3.0988 2.

5 and 4.4427 "10 !3 ) x 2 To determine the 98% confidence interval.681" (0. 2.02 " 4.18) At xi = 3.5 % 6.1960 " 10 ! 6 ) x 4 14 ! 2 # (0.4) 2 1 % 14 (220630000 ! 14 " 3721.5 % 6. ( )Y X ' 0.95) (4.4 2 ) !3 3 2 # (15274.98 # (7595.5 "10 3 ) & 2.5 " 10 3 ) & 2.98 # (7595.5 "10 ) ) % 14 (220630000 ! 14 " 3721.4) 2 1 % 14 (220630000 ! 14 " 3721.4) 2 1 % 14 (220630000 ! 14 " 3721.4 2 ) .681" (0.5.5 "10 3 ) (1.82 \$ 25291.02 " 3.98 # (7595.4427 "10 !3 " (4.5.5 "10 3 ! 3721.5 "10 3 ) 2 ) (3.4427 "10 !3 " (3.05 \$ 42057.5 % 6.5 "10 3 ! 3721.12 = 2.18 \$ 32600.02 " 2.5 "10 3 ! 3721.681" (0.681 from Table A.98 # (7595.99.5 "10 3 ! 3721.5 % 6.4) 2 1 & 2. ( )Y X ' 0. 3.5 " 10 3 ) 2 ) # (27313.5. we obtain. and with t0. let us use the following selected values of xi = 1. ( )Y X ' 0.4427 "10 " (1.5.(b) s 2Y sY (c) X X # 2.3523 " 10 !6 4 x # (0.3.681" (0.5.4427 "10 !3 " (2.82) At xi = 4.5 " 10 3 ) 2 ) (2.53 \$ 17976.5.4 2 ) # (19999.4 2 ) # (24730. At xi = 1.02 "1.5 " 10 3 ) & 2. ( )Y X ' 0.5.47) At xi = 2.

5423) P(Y > 30.25) = 0.40 .4427 # 10 \$3 " (3500) 2 ! 5423 Assuming Y is N(2.8665 x 104.02x3500 = 2.%(0.000) = 1 .%[(30000-28665) / 5423] = 1 .8665x10-4 sY X ! 0.5 + 6.(d) If x = 3500 E(Y X) = 7595.

.

1) 3 2 3 2 3 2 = ( ) (0.9) = P(p=0.01 (c) P(p=0.9) P(p=0.2652 = 0.1701 + 0.1"0.16 P’’(p=0.001 = 0.828 (d) Let A = event of two survival and one failure in three tests. P(pi!A) = P(A!pi) P(pi) / P(A) P(A) = P(A!p = 0. P(p=0.01 i = 0.9"0.35 + 0.001 = 0.09375 / 0.5) P(p=0.1)2"0.7 + 0.05 P(survival of the structure) = P(S) = P(S!p=0. P’’(p=0.25.1) P(p=0.01 = 0.7 / 0.05 = 0.05 = 0.10) = 0.1!A) = 0.7 + ( )(0.5) = 0.9) P(p=0.9"0.10"0.9!A) = 0.05 / 0.5!A) = 0.576 + 0.9) / P(E) = 0.5"0.25 / 0.5) P(p=0.5"0.25 + ( )(0.70.175 +0.2652 So.9!E) = #p i P’’(p=pi) = 0.9"0.64 P(P=0.16 + 0. So. P(p=0.9)"P(p=0.5"0. P(p=0.00135 = 0.5) + P(S!p=0.83 Similarly.9) + P(S!p=0.5"0.5)2"0. P’’(p=0. P(P=0.5"0.1"0.5) = 0.1) P(p=0.76 = 0.25 + 0.2652 = 0.9) + P(A!p=0.2652 = 0.080 + 0.1) = 0.00135 / 0.752 .747 + 0.09375 + 0.9.9"0.5) + P(A!p=0.9) = 0.35 P(P=0.1 (a) Let p = Probability that the structure survive the proof test.9"0.83 + 0.9!E) = P(E!p=0.76 = 0.76 = 0.1"0.1701 / 0.10) =0.01 E(P!A) = 0.64 + 0.1"0.76 (b) Let E = One structure survive the proof test.9)2"0.

4) + P(E2!p=0.276 So P’’(p=0.4) = P(p=0. P(E2) = P(E2!p=0.128) = 0.25 / 0.6) = 0.096"0.4) P(p=0.6"0.2"0.0) = 0 E(p!E2) = 0.8) + P(E1!p=1.2 (a) Let E1 = The output will be acceptable in the first day.4) = P(p=0.533 (b) Let.4) + P(E1!p=0.30 = 0.6 + 0.25 / 0.8 + 0 = 0.30 = 0.25 + ( ) (0.4"0.260 P’’(p=0.50 P’’(p=0.4!E2) = 3"0.4"0.25 + ( ) (0.25 + ( ) (1.128"0.8) + P(E2!p=1.4!E1) = 0.276 = 0.096 + 0.8"0.4"0.25 + 0.6)2"0.0)2"0"0.25 / 0.8)2"0. P(pi! E1 ) = P(E1!pi)"P(pi) / P( E1 ) P(E1) = P(E1!p=0.276 = 0.2"0.6"0.6) P(p=0.25 + 0. E2 = one unacceptable out of three trials.25 / 0.276 = 0.0) P(p=1.333 P’’(p=0.0) = 0 E(p!E1) = 0.0) 3 2 3 2 3 2 3 2 = ( ) (0.25 (0.6) P(p=0.6) + P(E1!p=0.8) = 3"0.8) = 0.2"0.30 So P’’(p=0.25 / 0.167 P’’(p=1.25 + 0"1.4)2"0.26"0.9.348 P’’(p=1.0 = 0.144"0.0) = 0.25 / 0.392 P’’(p=0.6) = 3"0.25 = 3"0.4"0.333 + 0.30 = 0.617 .6) + P(E2!p=0.6"0.8) P(p=0.4) P(p=0.348"0.6"0.144 + 0.167 = 0.392"0.0) P(p=1.4 + 0.5 + 0. Applying Bayes’ Theorem.8) P(p=0.

25 / 0.25 + (0.4) = P(p=0.0) P(p=1. .4!E3) = 0. P(E3) = P(E3!p=0.25 + 0 = 0.144 + 0.4) + P(E3!p=0.617 The posterior distribution will be same as shown in part (b) since the likelihood function of the observed events E2 and E3 are the same.25 + (0.8 + 0 = 0.4 + 0.26"0.144"0.092 = 0.096"0.0) = (0.(c) Let E3 = Out of three trial.8) + P(E3!p=1.096 + 0.348 P’’(p=1.6 + 0.8) P(p=0.25 / 0.348"0. first two are satisfactory and the last one is unsatisfactory.25 / 0.4) P(p=0.260 P’’(p=0.128) = 0.6) = 0.8) = 0.2"0.092 = 0.6)2"0.0) = 0 E(p!E3) = 0.25 (0.392 P’’(p=0.6) + P(E3!p=0.4)2"0.8)2"0.128"0.092 So P’’(p=0.4"0.6) P(p=0.6"0.092 = 0.392"0.

16 = 0.3 (c) P(HAHF! L A L F ) = P( L A L F ! HAHF) P(HALF) / P( L A L F ) Now using similar method as in (a).1"0.125 P(LALF! L A L F ) = 0.2 + 0.2"0.16 So.3.3 (b) P(HAHF! H A H F ) = P( H A H F !HAHF) P(HAHF) / P( H A H F ) = 0.3125 .25"0.16 = 0. P(HALF! L A L F F) = 0.3 + 0. P( L A L F ) = 0.3 P(HAHF) = P(HALF) = 0.2 (a) P( H A H F ) = P( H A H F ! HAHF) P(HAHF) + P( H A H F ! HALF) P(HALF) + P( H A H F ! LAHF) P(LAHF) + P( H A H F ! LALF) P(LALF) = 0.3 + 0.2 = 0.2 / 0.25"0.375 Similarly.1"0.1875 P(LAHF! L A L F ) = 0.4"0.3 / 0.2"0.10"0.2 + 0.3 = 0. P(HAHF! L A L F ) = 0.3 + 0.16 = 0.3"0.10"0.2"0.25"0.3"0.16 = 0.2 / 0.3 + 0.2 = 0.9.3 / 0.3 / 0. P(LAHF) = P(LALF) = 0.

352 Similarly.352"0.6 = 0. P(N # 2!m=0. N = Number of accurate measurements.4) P(m=0.2)0 = 0.667 = P’’(m=0.4) + P(X=2!m=0.896"0. 0.6 = 0.4 (a) P(X=2) = P(X=2!m=0.8) = P(N # 2!3.8"0.4) + P(N # 2!m=0.5 + 0.8) P(m=0.4) P(m=0.8) = 0.4) P(m=0. P(N # 2) = P(N # 2!m=0.4"0.4) = P(N # 2!3.4"0.4)3"(0.6)0 = 0.8) = ( )" (0.8) (b) Let.2 + ( )" (0.8! X=2) = 0.333 + 0.8) 3 2 3 3 3 2 3 3 P(N # 2!m=0.5 / 0.8) P(m=0.8"0.9.4! X=2) = 0.8)3"(0.4) = ( )" (0.896 Considering posterior distribution of m P(N # 2) = 0.5 = 0.8)2"0.4)2"0. 0.333 = P’’(m=0.6 P(m=0.6 + ( )" (0.714 .667 = 0.5 / 0.

P’’( ! = 4"E) = P’’( ! = 4) = 0.5 2 ) 5 4 .511 (b) f’’M(m) = fM(m" ! = 4) P’’( ! = 4) + fM(m" ! = 5) P’’( ! = 5) 1 m 1 m = m \$ 2 ( 5 )2 m \$ 2 ( 4 )2 e % 0 .01706 = 0.02085 dm2 where dm = small increment of m around m.01706 dm2 So.4 / 0.01706 = 0.02085%0.5 2 ) 4 4.5 \$ 2 ( e 52 1 5.01454%0.02044m # e 1 m \$ ( )2 2 5 0 2 0 ] +[-25%0. P( ! = 5) = 0.6) dm2 = 0. and P(E" ! = 5) = { 1 4.5 P( ! = 4) = 0. P(E) = (0.02085%0.6 / 0.4 + 0.02044 # e 1 m \$ ( )2 2 5 ] 2 0 dm .03056 # e = 0.02044m # e 1 m \$ ( )2 2 4 2 dm + 0 = [-16%0.03056m # e + 0.9.4.5 \$ ( P(E" ! = 4) = { 2 e 2 4 1 5.09675 1 m \$ ( )2 2 4 1 m \$ ( )2 2 5 & 0. P(E) = P(E" ! = 4) P( ! = 4) + P(E" ! = 5) P( ! = 5) 1 4.2 \$ ( # dm} { 2 e 2 4 # dm} = 0.03056m # e 1 m \$ ( )2 2 4 2 (c) P(M<2) = & 0.511 42 52 = 0.2 # dm} { 5 .01454%0.2 2 ) 4 5. 489 + e % 0.2 \$ 2 ( 5 ) 2 e # dm} 52 So.489 P’’( ! = 5) = 0.6 (a) Let E = Observed test results of bending capacity for the two tests.

71 So. 2% ! %3 .84 ! 2 (1 ' 1 ! )(1 ' = 0.9.0 . k = 14.)de} { (1. 3 4 )k ! 2 (1 ' 2 or.5 Let F = errors found from two measurements.5 de2 = de 81 81 P’’( ! =2"F) = P’’( ! =2) = 0 and P’’( ! =3) = 1 E( ! "F) = 2#0 + 3#1 = 3 cm (b) f’’( ! ) = k \$ L( ! ) \$ f’( ! ) where f’’( ! ) = 1. else where To find the constant k. f’’( ! ) = 58.6 (a) P( ! =2) = P( ! =3) = 0. P(F) = P(F" ! =2) P( ! =2) + P(F" ! =3) P( ! =3) 2 1 2 2 (1. 4k [' 1 ! * 1 ! )(1 ' 3 2! 2 ' 2 ! ) & d! ( 1. elsewhere 2 ! ) .)de} = de 3 3 3 3 81 P(F" ! =2) = { Where de = small increment of e So.)de} = 0 2 2 2 2 2 1 2 2 8 2 P(F" ! =3) = { (1. 2% ! %3 L( ! ) = { 2 ! (1- 1 )} { ! 2 ! (1- 2 ! )} So. f’’( ! ) = k & 4 ! 2 (1 ' 1 ! )(1 ' 2 ! ) &1 . P(F) = 0 + 8 4 2 #0. 2% ! %3 = 0.0 2 3! 3 ] 3 = 1.0 2 or.)de} { (1.

!2 2 .E( ! "F) = 3 3 % !f ' ' (! )d! \$ % 2 2 58.85 [ln(! ) & (1 # 3 ! # 1 ! )(1 # 2 ! )d! 1 3 ] = 2.84 ! = 58.61 cm.

0 ''( k " 4.271 #12 k & ) 12 e d! " 1. 20 ! 0.7 Since.9. where f’’(! ) = kL(! ) f’(! ) L(! ) " e # ! 12 ( ! 12 1! )1 . f ' (! ) " 0. 12 = 0.1082e 12 .5 \$ ! \$ 20 .5 \$ ! \$ 20 = 0. elsewhere 0.271 #12 e " 0.79 0.271 ! . elsewhere So. elsewhere To determine the constant k. 0.271 ! . 0. f ' ' (! ) " 4.5 \$ ! \$ 20 = 0.79 % ! ! # 0. f’’(! ) = k & e # ! 12 % ! 12 % 0.5 So the posterior distribution of ! is.

0 " 0.9 " p " 1.9 ! 3 & k ! 7 \$ p dp % 1.9 " p " 1.9 0. f’’(p) = 0.0 ! or.508 p dp # & 10.9 &! *+. =0.9 0.9) = 0.)- So.508 p3 .9 1.663 p dp !'( & ) h i .8 (a) P(p > 0. p 4 0.523 k[ ] # 7k [ ] % 1.9 4 4 & 0. =0.0 '' 3 4 0 4 0.3 (b) f’’(p) = k ! L(p) ! f’(p) 3 3 L(p) = ( ) p3 ! (1-p)0 = p3 0 " p " 0.9 1 So.0 0 0.663 p3 .0 1 &3 % \$ # To find k. = 10. elsewhere &" k ! p dp # 3 0 1.9) = 0.0 0. 0.9.9 p 4 1. = k ! p3 ! 7. elsewhere (c) E ''( p ) % 0 " p " 0. f’’(p) = k ! p3 ! 1/3.0 ( k % 1.7 and P(p > 0.

933 p 5 0.663[ ] 5 0 5 0.9 .508[ = 0.0 ] ! 10.= 0.9 p 5 1.

P(! =15) = 1/3 (a) Let X = Number of occurrences of fire in the next year.19 \$ # 0.0079 + 0.0592 = 0.0799 .81 20! 20! = 0.19 0.0731 (b) P’’(! =15) = P( X & 20! & 15) ' P(! & 15) P( X & 20) & 0.9 P(! =20) = 2/3.9.0592 & 0.0731 Similarly.0731 (c) P(X=20) = P(X=20"! =15) P’’(! =15) + P(X=20"! =20) P’’(! =20) e %15 (15) 20 e %20 (20) 20 = # 0.0139 + 0.0139 & 0.0720 = 0. P(X=20) = P(X=20"! =15) P(! =15) + P(X=20"! =20) P(! =20) = e %15 (15) 20 1 e %20 (20) 20 2 # \$ # 20! 3 20! 3 = 0. P’’(! =20) = 0.81 0.

240 (c) The updated distribution of ! as prior distribution in this problem: P '(! " 1/ 5) " 0.452 " 0.548 " P "(! " 1/ 5) " \$3 P(# ) 8.10 ! is the parameter to be updated with the prior distribution of: P '(! " 1 5) " 1/ 3 P '(! " 1 10) " 2 / 3 (a) Let # =accidents were reported on days 2 and 5.452 Let # = after the second accident in 10th day no accident was observed until the 15th day. P(# ) " P(T ' 10) " 1 \$ P(T ( 10) " 1 \$ P(T ( 10 | ! " 1/ 5) P "(! " 1/ 5) \$ P(T ( 10 | ! " 1/10) P "(! " 1/10) " 1 \$ (1 \$ e \$ 10 5 ) & 0.9.548 " 0.949 & 10 Combining the observation with the prior distribution of ! with Bayesian’s theorem. the posterior distribution of ! can be calculated as follows: 1 2 1 3 1 & exp( \$ ) & & exp( \$ ) & P(# | ! " 1/ 5) P '(! " 1/ 5) 5 5 5 5 3 " 0.452 (b) Let # =no accidents will occur for at least 10 days after second accident.949 & 10 P "(! " 1/10) " 1 \$ 0.548 \$ (1 \$ e \$ 10 10 ) & 0.548 P '(! " 1/10) " 0. The probability to observe Z is: . The probability to observe # is P(# ) " P(# | ! " 1/ 5) P '(! " 1/ 5) % P(# | ! " 1/10) P '(! " 1/10) 1 2 1 3 1 1 2 1 3 2 " & exp( \$ ) & & exp( \$ ) & % & exp( \$ ) & & exp( \$ ) & 5 5 5 5 3 10 10 10 10 3 \$3 " 8.

274 \$ 0.P(! ) \$ P(T % 5) \$ [1 # P(T & 5 | " \$ 1/ 5)]P '(" \$ 1/ 5) ' [ P (T & 5 | " \$ 1/10)]P '(" \$ 1/10) # 5 \$ [1 # (1 # e 5 )] ( 0.449 ( 1 nc ! P "(" \$ ) \$ 5 P( n \$ nc ) 0.1nc 1 nc ! P "(" \$ ) \$ 10 P( n \$ nc ) 0.476 P "(" \$ 1 10) \$ 1 # 0.202 ' 0.076 . nc \$ 2 . (e) Non-informative prior for " is used here.074 ( 2 nc 1nc ' 0.385 ( 1 5 Set P "(" \$ ) \$ P "(" \$ 1 ) 10 We get nc \$ 1.476 Combining the observation with the prior distribution of " with Bayesian’s theorem.575 (d) 1 1 1 1 P( n \$ nc ) \$ P ( n \$ nc | " \$ ) P(" \$ ) ' P( n \$ nc | " \$ ) P(" \$ ) 5 5 10 10 nc nc 2 1 \$ ( e #2 ( 0.548 ' [1 # (1 # e # 5 10 )] ( 0.548 ' ( e #1 ( 0.452 \$ 0.424 \$ 0.424 P( Z ) 0. Likelihood function is: L(" ) \$ " e #2 " " e #3" \$ " 2 e #5" Posterior distribution: .156 ( nc ! nc ! 0.2nc 0. Since nc is an integer. the posterior distribution of " can be calculated as follows: P "(" \$ 1 5) \$ [1 # P(T & 5 | " \$ 1/ 5)]P '(" \$ 1/ 5) 0.425 nc ! nc ! \$ 0.202 \$ \$ 0.

5! 2 e "5! .f "(! ) # L(! ) \$ ! 2 e "5! As ( %& 0 ! 2e "5! d! \$ 1 %& '(3) 2 \$ (5! ) 2 e " (5! )d(5! ) \$ ( 25 0 25 25 Thus the posterior distribution is: f "(! ) \$ 12.

25) . and the sample data.371 k "/ v " (b) P( X " 0) " ) \$% ") \$% 0 0 \$% P( X " 0 | m) f '( m)dm (0.209 1.4m)0 &0.225(1. k ' " 6.625 (a) It follows that the posterior distribution of M is also gamma distributed.25&1 &1.225 k " 7. It’s convenient to prescribe a conjugate prior to the Poisson process.25 .4 v' k '/ v ' .11 M is the parameter to be updated. v ' " 0.6 " 1. Thus.4 m 0 " 0. From the relationship given in Table 9.225m dm ' 'e ((7.1 between the prior and posterior statistics. the posterior distribution parameter of M can be estimated as: k " " k '\$ 1 " 7. From the information given in the problem.918 v " 1. the mean and variance of the gamma distribution of M is: E '( ! ) " k' k '/ v '2 " 10 # '( ! ) " " 0.4 m v "( v " m) k " &1 & mv " 'e ' ' e dm 0! ((k ") ") e &0.9.25 E "( M ) " " " 5.225m )7.25 v " " v '\$ 0.225 # '( M ) " k "/ v "2 " 0.

\$ !' ) = N(0.9.84.08 2 = 0. 0.94 0. \$ n ) = N(0.84.834 and.2) L( ! ) = N(0.0743 0. 9.75 # 0.91 # 0.08) = N ( x.8.84. 0.89 # 0.834 ) " ' (1.81) " 0.834.2 2 # 0.95 & 0.8 % (0.13.08 2 = 0. 0. \$ n ) From Equation 9.2) 2 # 0.95) = ' ( 0.2 2 # 0.0743) (c) P( ! < 0. f’’( ! ) = N ( x.84 4 (a) From Equation 8.08) (b) f’( ! ) = N( ! !' .15. ! !'' " x(\$ !' ) 2 # ! !' \$ 2 / n (\$ !' ) 2 # \$ 2 / n " 0.0743 .84 % (0.56) " 0. 0. 0.16) Sample mean x " 1 (0. f’’( ! ) = N(0.12 Let X = compression index X is N( ! .2 2 % 0.14.08) 2 0.08 2 So. 0. \$! " '' (\$ !' ) 2 (\$ 2 / n) (\$ !' ) 2 # (\$ 2 / n) " 0.16/2) = N(0.

4.33) min.58) min. L( ! ) = N (60.47 2 % 3.13 (a) T is N( ! .58 min .4.47 2 \$ 3. f’( ! ) = N(65. 10) Sample mean = t = 65 min. 102 \$ 2.9.47 2 \$ 3. (b) We know.17.16 2 f’’( ! T) = N(61. 10. f ' ' ( ! ) " N (t . # n ) " N (65.47 2 " 61.47) min.7 min .162 \$ 60 % 4. From Equation 8. 2. 10 5 ) " N (65.582 ) = N(61.14 and 8.7. (c) From Equation 9. 10 10 ) = N(60.4. n=5 So the posterior distribution of ! . P(T > 80) = 1 & ' ( 80 & 61.162 " 2.7. 3.0384 10.162 4.47) min.77) " 0. ! !'' " T #! " '' T 65 % 3.16) min.7 ) " 1 & ' (1. fT(t) = N(61.7. 4.33 .15.

224’ (b) The prior distribution of # can be modeled as N(32o.726’ % 1. N(32. 0.03' ! 0.0333 Hence.0167 % 0. .0204 2 " 0. 32 & 0.0121o = 32o0. 0.0204 2 " 32.05 o 6 ) or. 9. 0.044’ Applying Eqs.0333o).0204 " 0.2 6 \$1 s# ! 3.14 (a) The sample mean # ! 2 s# ! 1 (32 o 04'"31o 59'"32 o 01'"32 o 05'"31o 57'"32 o 00' ) ! 32 o 01' 6 1 {(3' ) 2 " (2' ) 2 " 0 " (4' ) 2 " (4' ) 2 " (1' ) 2 } ! 9. the Bayesian estimate of the elevation is. 2’) or N(32o.0167o.1.044' 2 2 0.044’ # is 32o0.0204 2 ' 0.9.0333 2 #''! 0.0333 2 ! 0.726’ and the corresponding standard error is.0167 & 0. ( #'' ! 0.15.0333 2 = 32.0204o) The value of the angle would be 32.14 and 9.01740 o or 1.0204o or 32o01’ % 1.05 o The actual value of the angle is N (32 o 01' .

20 % " 2 \$ 2.15 and considering First and Second measurements.14 and 9. 9.8944" " 2 \$ (2" ) 2 Now consider L’’ with the third measurement and apply Eq.18.20.8944" ) 2 L' ' ' # # 2.1616km (3" ) 2 \$ (0.18 % (0. " L'' # " 2 % (2" ) 2 # 0. ") Second measurement ! L=N(2.16km " 2 \$ (2" ) 2 and. 3") Applying Eqs. First measurement ! L=N(2.8944" ) 2 .14 again.9. L' ' # 2. 2") and Third measurement ! L=N(2.16 % (3" ) 2 \$ 2. 2.15.15 % (2" ) 2 # 2. 9.15 Assume.

1.4 " 32. for a binomial basic random variable.6 So E’’(p) = 3.6 + 12 – 1 = 32.16 Assume the prior distribution of p to be a Beta distribution.4 + 1 = 3.(1) q'" r ' and Var ' ( p) ! a' r ' ! 0.6 " 1) . we get. then.4 # 32.4 = 21.4 and r’ = 9 # 2.06 2 ---------(2) 2 (q'" r ' ) (q '" r '"1) From Equations 1 and 2.4 v’’ = v’ + n-x = 21.4 / (3. q’ = 2.9.6) 2 (3.6 From Table 9. q’’ = q’ + x = 2.1 ----------------------------.6) = 0. E ' ( p) ! q' ! 0.6 ! 0.4 + 32.0945 Var ' ' ( p) ! 3.4 " 32.00231 (3.

) " N (2.3542 '' (b) From equation 9.5.6 (a) From the observation we get t " N( 2 % 3 0.673) P(T ' 1) " P ( T & 2.673 .3542 '' 0. 0.9.17 ! is the parameter to be updated with the prior distribution parameter !!' " 3 # !' " 0.3052 ) " N (2. 0.305 0.62 % 0.354) 2 2 From the relationship given in Table 9. # 2 % # !'' ) " N (2.629 ' ) " ( ( &2.629 0.2 \$ 3 " 0.0.17 the distribution of T is 2 T " N ( !!" .629.421) " 0.629.5 !! " " 2.3542 \$ 0.673 0.629 1 & 2.62 % 0.774% 0.62 #! " " 0.62 \$ 2.0.354 2 \$ 3 % 0.5 .62 % 0.1.

q ' ! r ' ! 2 (b) Upon the observation the posterior distribution of p can be updated in terms of p and q as: q " ! q '" 2 ! 2 " 2 ! 4 r " ! r '" n # x ! 2 " 1 # 2 ! 1 f "( p ) ! \$(5) p3 ! 4 p3 \$(4) \$(1) (c) Let % =win the next two bids 1 1 0 0 p (% ) ! & P(% | p ) f "( p )dp ! & p 2 !4 p 3dp ! 2 3 . Thus.9. It’s suitable to set a conjugate prior for the problem. Thus.05 2 q '" r ' ( q ' " r ') ( q ' " r ' " 1) .18 (a) The parameter needs to be updated is p. we can get E '( p ) ! q' q'r' ! 0. With the information presented in the problem. suppose p is Beta distributed with parameter p’ and q’.5 Var '( p ) ! ! 0.

we assume Gamma distribution as prior distribution of !.5 k’’ = k’ + n = 25 + 2 =27 E’’(!) = k’’/v’’ = 27/52.193 0. v' Var ' (! ) # k ' 0 . From Table 9.5 = 0.514 Var’’(!) = k’’/v’’2 = 27/52.9.5. v’’ = v’ + %x i # 50 \$ (1 \$ 1.099 # 0.5 # # (0.796 x 10-3 &’’(!) = 0.52 = 9. E ' (! ) # k' # 0.099 ' !'' # 0. for an exponential basic random variable.1.5 " 0.5) # 52. v’ = 50 and.514 .19 Using conjugate distributions.2) 2 2 v' v' So. k’ = 25 Now.

9. thus f "(! ) & L(! ) \$ ! 4 e "24 ! (1 " e "6 ! ) ) '( 0 ! 4e "24! (1 " e "6! )d! \$ 1/ 23415 thus f "(! ) \$ 23415! 4 e "24 ! (1 " e "6 ! ) . The probability of crack length larger than 4 is: P( X # 4) \$ 1 " P ( X % 4) \$ e "4 ! The probability of crack length smaller than 6 is: P( X % 6) \$ 1 " e "6 ! Likelihood function is L(! ) \$ ! e "3! !! e "5! !(1 " e "6 ! )!! e "4! !e "4 ! !! e "8! \$ ! 4 e "24 ! (1 " e "6! ) non-informative prior is used.20 ! is the parameter to be updated. Let X denote the crack length.

344 where P( v # 120) " 1 ! P (v ( 120) " 1 ! P( T ! 161.653 2 152 ' 7. 152 % 6.344 " 0.209 (3) The probability that the house will be damaged in the next 5 years is P( v # 120.96 1.854 152 2 % 7.143 & \$" " 22 15 2 % 7.1'5 " 0. & \$" % & 2 ) " N (164.209) P( v # 220) " 1 ! P( v ( 220) " 1 ! P( T ! 164. thus & \$ " 1.876 ' 0.854 ) ) " 1 ! * (3.9.756 P( x # 0) " 1 ! P( x " 0) " 1 ! e !0.656 .402) " 0.876 35.854.1432 ) " N (164. The probability of x occurrences during time t (in years) is P( X " x ) " (0.157) " 0.756 35.393 " 0.1t ) x !0.17.1t e x! The probability that the tornado hits the town during the next 5 years is P( X # 0) " 1 ! P( X " 0) " 1 ! e !0.21 (a) The occurrence rate of the tornado is estimated from the historical record as 0.376 ) ) " 1 ! * ( !1.6532 2 " 164.393 The probability that the house will not be damaged is 1 ! 0.96 & \$' The updated the distribution parameter of mean maximum wind speed is \$\$" " 160 ' 152 % 175 ' 7.854.96 . the updated the distribution of maximum wind speed is 2 v " N ( \$\$" .209 16.6532 " 6. x # 0) " P ( v # 220) P( x # 0) " 0.033% 16.376 120 ! 161.393 The prior distribution parameter of mean maximum wind speed is: \$\$' " 145 % 175 " 160 2 \$\$' ! 145 \$\$' ! 145 15 ' " " 7.653 2 From equation 9.854 220 ! 164.1/year.653 " 1.16.5 " 0.

22 h is the parameter to be updated.003h 2 ! 3.1428 h .5h!0.5712 0 Where 10 10 4 4 f ( x ) ! " f ( x | h ) f "( h )dh ! " 1 !0.003h 2 ! dh ! 0. (a) 10 10 0 0 E ( x ) ! " E ( x | h ) f ( h )dh ! " 0.003h 2 ! ! 0.003h h 10 1 Since " 0.5h h (b) (i) the range of h is 4<h<10 upon the observation (ii) likelihood: 1 h L(h | x ! 4) ! p( x ! 4 | h ) ! Prior: f '( h ) ! 0.1428 % 4 ! 0.0238h 0.132 (iii) The probability that the hazardous zone will not exceed 4km in the next accident is: 4 P( X \$ 4) ! " f ( x )dx ! 0.132 .003h ! 0.003h 2 Thus the posterior is 1 f "( h ) # 0.9.0238hdh ! 0.75 where h h 0 0 E ( x | h ) ! " xf ( x | h )dx ! " x dh ! 0. the posterior is: 4 h 1 f "( h ) ! !0.

052 4 & 0. \$ 2 " \$ "2! ) # N (0.96 \$ & sL # 0.85 % 0.85 2 . From the information in the problem.83) P( X 2 * 0.23 (a) Let x denote the fraction of grouted length.88.7782 # 0.85 ' 0.05 1.9.026 0.96 # !!' % 0.85 % 0.0654 2 The probability that the minimum grouted length is smaller than 0.05 4 The posterior covariance is 0. \$ ! # ' !!' % 0. . the variance of fraction of length grouted to the constant and can be approximated by the observation data.85 is P( X ( 0.88 0.062 ' 0.062 2 " 0.75 1.88 ) # 0.052 4 # 0. the mean grouted length is normally distributed.88=3% X # N ( !!" . we have 0.026/0.95 !! # # 0.75 # 0.96 \$ !' 0.75 " 0.83) # 1 % 0.0.062 " 0.062 2 " 0. Assumptions: the grouted length is normally distributed. ' Thus.395 So the second requirement is more stringent.06 !!" # \$ !" # 0.85) # ) ( 0.05 4 0.83 is 1 % P( X 1 * 0.258 0.8 # 1.89 ' 0.0654) (b) The probability the sample average grouted length is lessen than 0.

9. tn | !T . n 2" T2 2&" T . 1989) f '( !T . 2" T2 . t2 . Edward Arnold.2 2 density \$ % So " T2 " * Inv ) 2 2 ( n ) 1. 2&" T 1 % 1 % where s 2 * 1 n (ti ) t ) 2 0 n ) 1 i *1 (3) Joint posterior is: ' ( n ) 1) s 2 / n ( t ) ! ) 2 ( f "( !T . s 2 ) . (4) Marginal posterior of " T2 f "(" T2 ) * 1 f "( !T . \$ ' ( n ) 1) s 2 / n ( t ) ! ) 2 ( exp n +) . 2&" T 2" T2 . " T2 ) # 1 " T2 (2) Likelihood p(t1 . d !T 2" T2 . 2" T2 ... Bayesian Statistics: An Introduction. - Which is a scaled inverse. " T2 ) # " T) n )2 exp + ) . ) n )1 2 ' ( n ) 1) s 2 ( exp + ) 2" T2 ... Lee. " T2 )d !T # 1" ) n )2 T # \$" T2 % ' ( n ) 1) s 2 / n ( y ) ! ) 2 ( exp + ) . 2&" T 2&" T 1 * * \$ ' 0 (ti ) !T ) 2 ( exp + ) .24 (1) Prior Non-informative prior is used as (P. 2" T2 .M. " T ) ' (t1 ) !T ) 2 ( ' (t2 ) !T ) 2 ( ' (tn ) !T ) 2 ( 1 1 * exp + ) " exp + ) "!" exp + ) 2" T2 .! .

Var ( !T" ) % 0" 1 2 " T 4 / 4. From the information available in the problem.(5) Marginal posterior of !T Similarly. S % (n # 1) s 2 % (5 # 1) / 102 % 400 !T" % tn #1 ( t .998 4#2 % Inv # 2 2 ( n # 1. " T# n #2 exp * # + d" T 2" T2 . only the first set of data with 5 flight time has both sample mean and sample standard deviation.472) E ( !T" ) % 65 .4722 % 39. the posterior of !T can be written as f "( !T ) % . #n / 2 #n / 2 which is the tn #1 ( t . z ( n #2) 2 exp( # z )dz 0 ) &. s % 10 . t % 65 . s 2 ) % Inv # 2 2 (4. s 2 n ) % t4 (65. s 2 n ) density. \$ ) A#n 2 . . 4.4#2 . A % ( n # 1) s 2 ( n( !T # t ) 2 (6) We need both sample mean and sample standard deviation in this problem. Or it can be written as !T" # t s n % tn #1 Where z % A 2" 2 T . f "( !T . " T2 )d" T2 & ( n # 1) s 2 ( n ( t # ! ) 2 ' 2 ) .( n # 1) s 2 ( n ( !T # t ) 2 '& n( !T # t ) 2 ' ) *1 ( ( n # 1) s 2 +. So only this set of data will be used. n % 5 .100) " 4 / 100 % 200 E &*0" T2 1 '+ % .

2) 2 (4 . 4) ." 2 \$ 42 Var %(!# T2 " &) ' \$ 100 ' * + (4 .

! " 0.512 .1 % .368 % 0.73. E (Y | x ) " 0.+ 1.368 if the value of the basic scatter \$ 2 is equal to 1.25 Let x denotes the rated value. and y denotes the actual value.167 From equation 9. Based on 9.536 ) " 0.751 . \$ 2 " 1.24 to 9.0169(24 * 24.167) 2 ) ' 2 + .512 % 0.73.34 1.368 P(Y 4 18 | x " 24) " 5 ( 18 * 18.1 % .9.536 Var (Y | x " 4) " 3.167) 2 " 3. # " 0. the variance of Y at x " 24 becomes 1.385 3. s x2 " 34.73 .751 + 24 " 18.27.34 as Var (Y | x ) " 6 *1 & 1 ( 6 ( x * 24.512 % 0.732 " 3.26. Then P(Y 4 18 | x " 24) " 5 ( 18 * 18.0169( x * 24.751x The variance can be calculated by equation 9.167) / 6 * 3 2 6 0 6 * 1 34.536 ) " 0.732 .167 1 3 E (Y | x " 24) " 0.368 % 0.

10) 0 (0.90) 20 ( %% ""(0.05 The welds should not be accepted.10) From trial and error.90 x (0. AOQL = 0. and r = 1 ' 20 \$ ' 20 \$ ""(0.10) ! 0.05 For n = 20. n = 25 and r = 1 So. and using Table of Cumulative Binomial Probabilities.10. (c) AOQ = pg(p) Here. r = 5.0332 . n = 105.2 g(0.90) 19 &1# &0# g(0.10) 1 (0.97) n + x * 0.1.2702 = 0.05 and from part (a) g(0.25 .1216 + 0.1 (a) Apply Equation 10. 1 AOQ = p ' 25 \$ )% x " p x *0 & # x (1 + p ) 25+ x * p .90) n + x * 0.10) = r ' n\$ x *0 & # ) %% x ""(0.03) x (0. (1 + p ) 25 ( p 2 .10) = r 'n\$ x *0 & # ) %% x ""(0. From the graph. (b) g(0.10) = %% = 0.(1 + p ) 24 This is plotted in Fig 10.

10.1 AOQ Curve .Fig.

2 P(SO2 < 0. to achieve a reliability of 0.10.90) .05) # 28.e.1 unit) ! 0.9 with 95% confidence n# ln(0.4 " 29days ln(0.90 with 95% confidence i.

10.3
(a) Rn = 1 – C
Here,
n = 10 and R = 1 - 0.15 = 0.85
So,
C = 1 - Rn = 1 - (0.85)10 = 0.8031
(b) C = 0.99, R = 0.85, n "

ln(1 # C ) ln(0.01)
"
" 28.3 ! 29
ln R
ln(0.85)

10.4

n"

ln(1 ! C )
ln R

Here,
C = 0.95 and R = 0.99
So,

n"

ln(1 ! 0.95)
" 298
ln(0.99)

298 consecutive successful starts would be required to meet the standard.

10.5
g(0.05) =

r

)n&

x "0

( %

! '' x \$\$(0.05)

x

(0.95) n # x " 0.94

and,
g(0.20) =

r

)n&

x "0

( %

! '' x \$\$(0.20)

x

(0.80) n # x " 0.10

The values of n and r are to be determined by trial and error. Alternatively, from fig. 10.1e, n = 30 and r =
3 approximately satisfy the above two equations.

10.6
(a) p = fraction defective = 0.30
n = 5, r = 0
From Equation 10.2,

' 5\$

g(0.3) = %% ""(0.3) 0 (0.7) 5 ! (0.7) 5 ! 0.168
0

& #

= consumer risk

' 5\$

(b) g(0.1) = %% ""(0.1) 0 (0.9) 5 ! 0.59
0

& #

! P(rejection) = 1 – g(0.1) = 0.41

!"#\$%
%
&'(% !%)%*+,-./0+12%+3242%) P & X \$ L & a ( ! # &

%

L " &a

%

( ! #&

!!6 " !8"
( ! # &"!#56( ! "#"5\$ %
7

n
!"
L " &t
!!6 " !!7
"%)%9,:2.;0+12%+3242%)% P & X ' L & t ( ! ! " # &
( ! ! " #&
( ! ! " # &>#!=( ! "#"""\$<8 %
%
7
n

%
&?(% !%)%"#!"%)% # &

L " !8"
(
7

n
L " !!7
"%)%"#"5%)% ! " # &
(
7

&

or

!"

L " !8"
( n ! "!#86 %
7

or

n&

L " !!7
( ! !#=7 %
7

n

L " !8" " !#86
!
( L ! !!\$#>\$ %
L " !!7 !#=7

%

!%

%
%
%
%
&/(%

%
@+,;%AB32C%
% :)%>#6C%2,%A'40%:%)%7#"%':-%D%)%!!\$#>7%E?FGA>
H23:I%0J.'A3,:2%!"#!"%':-%!"#!!C%

n & L " !8"(
! "t "#!"C n "!
7

and

n & L " !!7(
! t "#"5C n "! %
7

%
KGA0+%'%A+3'EL':-L0++,+%M+,/0-.+0C%:%32%G,.:-%A,%?0%=%':-%AB0%/,++02M,:-3:I%N'E.0%,G%D%?0/,;02%!!\$#5<%E?FGA>%
,+%!!\$#8<%E?FGA>#%%O,%A'40%AB0%'N0+'I0%D%3:%,+-0+%A,%?0%G'3+%A,%?,AB%/,:2.;0+%':-%M+,-./0+%':-%D%)%!!\$#77%E?FGA>#%
%

!"#\$%
%
%
!%&%"#!"'% !%&%"#!"%
%
()%&%"#"*'%(+%&%"#!"%
%
,-.%/.012/.3%4)567.%428.%9%24'%
*

%

*

%

*

& ' (! ;! ( + : ( ' (! ;* : #
& ' (! ;"#<": ( ' (! ;"#!": #
)
n ) \$ (!
!
\$ (!
! %
(!
(!
% ' ;"#!": ( ' ;"#"*: "
% ' ; Pt : ( ' ; Pa : "
*

& !#*\$ ( ;(!#*\$: #
& *#>? #
)\$
)\$
!
! ) !"#== , !! %
% "#=\$ "
% ( !#*\$ ( ;(*#"?: "

%
%
%%%%%%,-.%@A//.46A9329B%+A7./)C7.%D/)@+2A9%3.D.@+2E.%F%24%
%
%

%

& (!
& (!
' (! ;* : #
' (! ;"#!": #
M ) ' \$' ; Pa : (
! ) ' \$' ;"#"*: (
!%
n "
!! "
%
%
&
!#*\$ #
% ) ' \$( *#"? ! ) ' ;(!#?=G: ) "#"G= %
!! "
%

" !!#-.!"#\$% % % !&%'%"#\$"(%!)%'%"#*"(%!%'%"#"+(%"%'%"#"+(%#%'%"#". " !#-. % "#". or n/ L ! "#*" . n % % % 2345607%)89%&:359%). n % &01% ! ! # / L ! "#*" . " "#"+ "#". or n/ L ! "#\$" . " "#"+ "#". % % "#".3%9<=&)630>(%0%'%?%&01%@%'%"#*+#% .% % #/ L ! "#\$" .