SPECIAL ATTRIBUTES OF

LPP
TWO PHASE METHOD
DUALITY IN LPP
SENSITIVITY ANALYSIS
TWO PHASE METHOD
ALTERNATIVE TO BIG-M METHOD
UTILITY
PHASE I
1. STANDARDISE THE PROBLEM
 FOR EACH ARTIFICIAL VARIABLE, ASSIGN COEFFICIENTS IN
THE OBJECTIVE FUNCTION
 IN MAX PROBLEM: ASSIGN –1
 IN MIN PROBLEM: ASSIGN 1
 FOR EVERY OTHER VARIABLE, ASSIGN COEFFICIENT OF ZERO

2. SOLVE BY SIMPLEX METHOD. IF THE OBJECTIVE FUNCTION

VALUE FOR OPTIMAL SOLUTION IS ZERO, MOVE TO PHASE II

SOLUTION TO LPPS WHEN SOME/ALL

CONSTRAINTS ARE
"≥" TYPE: TWO-PHASE METHOD
PHASE II

1. BEGIN WITH SIMPLEX TABLEAU OF PHASE I FINAL SOLUTION,

ELIMINATE ALL ENTRIES FOR ARTIFICIAL VARIABLES AND
REPLACE THE ZERO COEFFICIENTS OF DECISION VARIABLES BY
ORIGINAL COEFFICIENT VALUES

2. SOLVE THE MODIFIED PROBLEM BY SIMPLEX METHOD

SOLUTION TO LPPS WHEN SOME/ALL

CONSTRAINTS ARE
"≥" TYPE: TWO-PHASE METHOD
AN AGRICULTURAL RESEARCH INSTITUTE SUGGESTED TO A FARMER TO
SPREAD OUT AT LEAST 4,800 KG OF THE SPECIAL PHOSPHATE FERTILISER AND
NOT LESS THAN 7,200 KG OF A SPECIAL NITROGEN FERTILISER TO RAISE
PRODUCTIVITY OF CROPS IN HIS FIELDS. THERE ARE TWO SOURCES FOR
OBTAINING THESE – MIXTURES A AND B. BOTH OF THESE ARE AVAILABLE IN
BAGS WEIGHING 100 KG EACH AND THEY COST RS 40 AND RS 24,
RESPECTIVELY. MIXTURE A CONTAINS PHOSPHATE AND NITROGEN
EQUIVALENT OF 20 KG AND 80 KG, RESPECTIVELY, WHILE MIXTURE B
CONTAINS THESE INGREDIENTS EQUIVALENT OF 50 KG EACH.

FORMULATE THE ABOVE PROBLEM AS AN LPP.

USING SIMPLEX METHD, DETERMINE HOW MANY BAGS OF EACH TYPE THE
FARMER SHOULD BUY IN ORDER TO OBTAIN THE REQUIRED FERTILISER AT
MINIMUM COST.

EXAMPLE- TWO PHASE METHOD

 MINIMISE : Z = 40X1 + 24X2
SUBJECT TO;
20X1 + 50X2 ≥ 4800
80X1 + 50X2 ≥ 7200

X1 ; X2 ≥ 0

LPP FORMULATED
 MIN: Z = 0X1 + 0X2 + 0S1 + 0S2 + A1 + A2

SUBJECT TO;
20X1 + 50X2 - S1 + A1 = 4800
80X1 + 50X2 - S2 + A2 = 7200
X1 ; X 2 ; S1 ; S2 ; A 1 ; A 2 ≥ 0

STANDARDISED PROBLEM – PHASE 1

 BASIS X1 X2 S1 S2 A1 A2 bi bi /X2

A1 1 20 50** -1 0 1 0 4800 96

A2 1 80 50 0 -1 0 1 7200 144

Cj 0 0 0 0 1 1
Zj 100 100 -1 -1 1 1
Δj = Cj - Zj – 100 -100 1 1 0 0

INITIAL SOLUTION
 TESTING OPTIMALITY
TABLE 2
BASIS X1 X2 S1 S2 A2 bi bi /X1

X2 0 2/5 1 -1/50 0 0 96 240

A2 1 60** 0 1 -1 1 2400 40
Cj 0 0 0 0 1
Zj 60 0 1 -1 1

Δj = Cj - Zj - 60 0 -1 1 0
 TESTING OPTIMALITY
TABLE 3

BASIS X1 X2 S1 S2 bi

X2 0 0 1 -2/75 1/150 80
X1 0 1 0 1/60 -1/60 40
Cj 0 0 0 0
Zj 0 0 0 0

Δj = Cj - Zj 0 0 0 0

IF THE OBJECTIVE FUNCTION VALUE FOR OPTIMAL SOLUTION IS

ZERO, MOVE TO PHASE II
 PHASE 2: TESTING OPTIMALITY
TABLE 3
BASIS X1 X2 S1 S2 bi bi /S1

X2 24 0 1 -2/75 1/150 80 --
X1 40 1 0 1/60** -1/60 40 2400
Cj 40 24 0 0
Zj 40 24 2/75 -38/75
Δj = Cj - Zj 0 0 -2/75 38/75

AFTER REPLACING THE ORIGINAL COEFFICIENTS IN TABLE

3
 TESTING OPTIMALITY
TABLE 4
BASIS X1 X2 S1 S2 bi

X2 24 8/5 1 0 -1/50 144

S1 0 60 0 1 -1 2400
Cj 40 24 0 0
Zj 192/5 24 0 -12/25
Δj = Cj - Zj 8/5 0 0 12/25

• THE ABOVE SOLUTION IS OPTIMAL AS ALL Δj ARE EITHER

POSITIVE OR ZERO
• THE SOLUTION CAN BE READ AS FOLLOWS:
• OPTIMAL VALUE OF X1 = 0; X2 = 144, S1 = 2400
• OPTIMAL VALUE OF Z = 40(0) + 24(144) = 3456
• THIS IS THE SAME SOLUTION AS WAS OBTAINED USING
GRAPHICAL METHOD
• SINCE S1 = 2400, CONSTRAINT ON PHOSPHATE FERTILISER IS
 ALTERNATIVE TO BIG-M METHOD
 SOLUTION IS DIVIDED INTO TWO PHASES
 PHASE 1 HELPS TO DRIVE OUT THE ARTIFICIAL VARIABLES
 IF PHASE 1 IS NOT COMPLETED, THE SOLUTION IS
INFEASIBLE
 PHASE 2 HELPS TO DETERMINE THE OPTIMAL SOLUTION

UTILITY OF TWO PHASE METHOD

CONCEPT OF DUALITY
HOW TO WRITE A DUAL

UTILITY
 EVERY LPP HAS A DUAL, THE GIVEN LPP IS PRIMAL

 IF PRIMAL IS A MAX PROBLEM THE DUAL IS A MIN PROBLEM

 AN M-VARIABLE N-CONSTRAINT PRIMAL HAS AN N-VARIABLE M-CONSTRAINT

DUAL

 THE DUAL IS OBTAINED BY TRANSPOSING THE THREE MATRICES INVOLVED IN

THE PRIMAL PROBLEM

CONCEPT
 BEFORE WRITING DUAL FOR AN LPP, MAKE SURE
1. ALL VARIABLES ARE NON-NEGATIVE
 REPLACE EVERY UNRESTRICTED VARIABLE BY DIFFERENCE OF
TWO NON-NEGATIVE VARIABLES

2. ALL CONSTRAINTS ARE “≤” TYPE FOR A MAX PROBLEM/ “≥” TYPE FOR
A MIN PROBLEM
 IF A CONSTRAINT IS IN A DIRECTION OPPOSITE TO THE ONE
DESIRED, THEN MULTIPLY IT BOTH SIDES BY -1 AND REVERSE THE
DIRECTION OF INEQUALITY
 REPLACE A CONSTRAINT INVOLVING “=” SIGN WITH A PAIR OF
CONSTRAINTS WITH IDENTICAL LHS AND RHS VALUES: ONE WITH
“≥” SIGN AND THE OTHER WITH “≥” SIGN

PRE-REQUISITES
 Primal Dual

Max Min

No. of variables No. of constraints

No. of constraints No. of variables

≤ type constraint Non-negative variable

= type constraint Unrestricted variable

Unrestricted variable = type constraint

Objective function value for jth variable RHS constant for ith constraint

RHS constant for ith constraint Objective function value for jth variable

Coefficient (aij) for jth variable in the ith constraint Coefficient (aij) for ith variable in the jth constraint

PRIMAL-DUAL RELATIONSHIP
 PRIMAL
MAXIMISE Z = 22X1 + 30X2 + 32X3
SUBJECT TO 5X1 + 3X2 + 8X3 ≤ 70
2X1 – 8X2 + 9X3 ≤ 86
X1 , X2 , X3 ≥ 0
HERE BOTH CONDITIONS ARE SATISFIED.

 DUAL
MINIMISE G = 70Y1 + 86Y2
SUBJECT TO 5Y1 + 2Y2 ≥ 22
3Y1 – 8Y2 ≥ 30
8Y1 + 9Y2 ≥ 32
Y1, Y2 ≥ 0
 HERE X1 , X2 , AND X3 ARE PRIMAL VARIABLES AND Y1 AND Y2 ARE DUAL VARIABLES
 THE PRIMAL HAS 3 VARIABLES AND 2 CONSTRAINTS, WHILE THE DUAL HAS 2 VARIABLES
AND 3 CONSTRAINTS

WRITING THE DUAL: EXAMPLE 1

PRIMAL
MAXIMISE Z = 3X1 + 5X2 + 7X3
SUBJECT TO 2X1 + 4X2 + 3X3 ≤ 40
–4X1 + 5X2 – 3X3 ≥ 25
X1 + 2X2 + 5X3 = 15
X1, X2 ≥ 0, X3: UNRESTRICTED IN SIGN

WRITING THE DUAL: EXAMPLE 2

 TO WRITE THE DUAL, FIRST MULTIPLY SECOND CONSTRAINT BY -1;
REPLACE THE THIRD CONSTRAINT BY A PAIR OF INEQUALITIES IN OPPOSITE
DIRECTIONS; AND REPLACE THE UNRESTRICTED VARIABLE X3 = X4 – X5
THE PROBLEM BECOMES:
MAXIMISE Z = 3X1 + 5X2 + 7X4 – 7X5
SUBJECT TO 2X1 + 4X2 + 3X4 – 3X5 ≤ 40
4X1 – 5X2 –3 X4 + 3X5 ≤ – 25
X1 + 2X2 + 5X4 – 5X5 ≤ 15
– X1 – 2X2 – 5X4 + 5X5 ≤ – 15
X1, X2, X4, X5 ≥ 0

WRITING THE DUAL: EXAMPLE 2

WITH DUAL VARIABLES Y1, Y2, Y3 AND Y4,
MINIMISE G = 40Y1 – 25Y2 + 15Y3 – 15Y4
SUBJECT TO 2Y1 + 4Y2 + Y3 – Y4 ≥ 3
4Y1 – 5Y2 + 2Y3 – 2Y4 ≥ 5
3Y1 – 3Y2 + 5Y3 – 5Y4 ≥ 7
– 3Y1 + 3Y2 – 5Y3 + 5Y4 ≥ – 7
Y1, Y2, Y3 AND Y4 ≥ 0

WRITING THE DUAL: EXAMPLE 2

 NOW, LET Y3 – Y4 = Y5 AND COMBINING THE LAST TWO CONSTRAINTS AND
RE-WRITING THEM INSTEAD AS AN EQUATION, THE DUAL IS:
MINIMISE G = 40Y1 – 25Y2 + 15Y5
SUBJECT TO 2Y1 + 4Y2 + Y5 ≥ 3
4Y1 – 5Y2 + 2Y5 ≥ 5
3Y1 – 3Y2 + 5Y5 = 7
Y1, Y2 ≥ 0 AND Y5 UNRESTRICTED IN SIGN

 FOR AN UNRESTRICTED VARIABLE IN THE PRIMAL, A CONSTRAINT IN DUAL

INVOLVES AN EQUATION, WHILE FOR AN EQUATION IN THE PRIMAL, THERE
IS AN UNRESTRICTED VARIABLE

WRITING THE DUAL: EXAMPLE 2

 IF FEASIBLE SOLUTIONS EXIST FOR BOTH PRIMAL AND DUAL,
THE OBJECTIVE FUNCTION VALUES OF THEIR OPTIMAL
SOLUTIONS ARE EQUAL

 THE SOLUTION TO THE DUAL CAN BE READ FROM THE ∆J VALUES

OF THE SLACK/SURPLUS VARIABLES IN THE OPTIMAL SOLUTION
TABLEAU OF THE PRIMAL

 IF PRIMAL HAS UNBOUNDED SOLUTION, THE DUAL HAS NO

FEASIBLE SOLUTION

PRIMAL-DUAL RELATIONSHIP
 THE ∆J VALUES CORRESPONDING TO SLACK/SURPLUS VARIABLES
IN THE OPTIMAL SOLUTION INDICATE MARGINAL PROFITABILITY
(OR SHADOW PRICES) OF THE RESOURCES THEY REPRESENT

 A RESOURCE UNUSED FULLY HAS ZERO SHADOW PRICE

 A UNIT CHANGE IN THE AVAILABILITY OF A RESOURCE CHANGES

THE OBJECTIVE FUNCTION VALUE BY AN AMOUNT OF ITS
SHADOW PRICE

ECONOMIC INTERPRETATION OF DUAL

SENSITIVITY ANALYSIS
MEANING
RESOURCE (bi)
COST (Cij)
 RHS RANGING
 SHADOW PRICES ARE VALID ONLY OVER CERTAIN RANGES
 THE RANGE FOR EACH RESOURCE IS DETERMINED BY bi AND aiJ
VALUES IN THE OPTIMAL SOLUTION TABLEAU

 CHANGES IN OBJECTIVE FUNCTION COEFFICIENTS

 WITHIN CERTAIN LIMITS, SUCH CHANGES DO NOT INDUCE CHANGES
IN THE OPTIMAL MIX
 THE LIMITS ARE DETERMINED BY ∆J AND THE APPROPRIATE aiJ
VALUES IN THE OPTIMAL SOLUTION TABLEAU

SENSITIVITY ANALYSIS
 ECONOMIC INTERPRETATION

SIMPLEX TABLEAU: OPTIMAL SOLUTION

Basis x1 x2 x3 S1 S2 S3 bi

x2 10 1/3 1 0 1/3 -1/6 0 8

x3 8 2/3 0 1 -1/3 5/12 0 10

S3 0 -8/3 0 0 1/3 -17/12 1 18

Cj 5 10 8 0 0 0

Zj 26/3 10 8 2/3 5/3 0

Δj -11/3 0 0 -2/3 -5/3 0

 EACH HOUR IN FABRICATION IS WORTH RS 2/3 SO THAT AN INCREASE OF ONE HOUR WOULD
INCREASE PROFIT BY RS 2/3 AND A DECREASE OF ONE HOUR WOULD RESULT IN A DECREASE
OF RS 2/3
 SIMILARLY, EACH HOUR IN FINISHING IS WORTH RS 5/3
 THE MARGINAL PROFITABILITY OF PACKAGING HOURS IS 0 SINCE THERE IS UNUTILIZED
CAPACITY OF 18 HOURS. FURTHER ADDITION TO CAPACITY WILL NOT ADD TO PROFIT AND
REDUCING CAPACITY WILL ALSO NOT AFFECT IT
 RHS RANGING

bi aij bi/aij
 FABRICATION HOURS
8 1/3 24 Least positive (Closest to Zero)
10 -1/3 -30
Highest negative (Closest to Zero)
18 1/3 54

 RANGE: 60-24 = 36 TO 60-(-30)=90

 SHADOW PRICE OF RS. 2/3 IS VALID OVER
THE RANGE 36 TO 90 HOURS
 FINISHING HOURS bi aij bi/aij

8 -1/6 -48

10 5/12 24 Least positive ( Closest to Zero)

18 -17/12 -12.71 Highest negative (Closest to Zero)
 RANGE: 72-24 TO 72+12.71 OR 48 TO 84.71
 PACKAGING HOURS
 100-18=82 ; THEREFORE, RANGE IS 82 TO ∞
 CHANGES IN OBJECTIVE FUNCTION COEFFICIENTS
For x2 (a basic variable)
Δj -11/3 0 0 -2/3 -5/3 0 Range: 10-2 to 10+10
(x2) ij 1/3 1 0 1/3 -1/6 0 OR
Rs. 8 to Rs. 20
Ratio -11 0 - -2 10 -

Least negative Least positive

For x3 (a basic variable)

Δj -11/3 0 0 -2/3 -5/3 0
Range: 8-4 to 8+2
(X3) aij 2/3 0 1 -1/3 5/12 0
OR
Ratio -11/2 - 0 2 -4 - Rs. 4 to Rs. 10
Least positive Least negative

Range: -∞ to 5+11/3 For x1 (a non-basic variable)

OR
Rs -∞ to Rs 8.67