You are on page 1of 81
Chapter + Solving Linear Pagramning Pablo! the 4d-d lent Hethed . Simp Note: comer points will be called A, B, C, D, and E going clockwise from (0,2). b) Comer Point ¢) Comer Point d) Comer Point e) Corner Points A: x20, andx,=2 x, = 2, and x, +2) x+y =3, and x, x, 22, and x, =0 x, =0, and x, =0 (ey) Qi 4) (ty 2) (2) (ty aD) E and B are adjacent Aand Care adjacent Band D are adjacent Cand E are adjacent Dand A are adjacent A and B share x, =2 Band C share x, + x, =3 Cand D share Dand E share Eand A share =3 =2 MOOS> MONP> MOOW 4a-e a) Optimal Solution: (x,.x, ) = (2,2) and Profit = 10 Note: comer points will be called A, B, C, and D going clockwise from (0,3). b) Corresponding Constraint ‘Corner Point Satisfies the Corner Point Boundary Equations Equations ‘A: ©) %,=0 x + 2x, B: (22) X, + 2k; 2k, +3 C: 3,0) 2(3)40=6 0=0 ‘D: (0,0) ‘0=0 0=0 °) Corner Point Its Adjacent Corner Points a3) j—coyand 2) B: (2,2) (0,3) and (3,0) C: (3,0) (2,2) and (0,0) D: (0,0) (3,0) and (0,3) @ Corner Point (x,.%) Profit = 3x, #2x, (0,3) 6 (2,2) 10 (3,0) 9 (0,0) oO Optimal Solution: (x,.% ) = (2,2) and Profit = 10. 4A 9 Comer Point_| Profit = 3x, +2x Next Step D: 0.0) d Check points A and C. A:(0,3) 6 fove to C. C: (3,0) 9 Check point B. B: (2,2) 0 sm, 2b optimal + the next comer point is A, which has already been checked GAB @- Corer Point (A,,A, Profit = $1000A, + $2,000A, @.0) 0 (8,0) $8,000 (6,4) $14,000 (5,5) $15,000 (0,6.667) $13,333, Optimal Solution: (A,,A,) = (5,5) and profit = $15,000. by these oe twoi (0,0) -» &.2) © (6,4) (5,5) Gre) is act optimal sinec le, 6-4) and (8,0) ene adyncent, feasible and Letter. (go) is not optimal since (6,4) is adyaceat, Feasible ond better ‘ a (s,5) 65 (GA) is nero imal since sana rue and better. (s,s) is optimal Sine bath adjeccat CFs’s are worst The other possible paths: 0)2(e64) 4 (3,5). 4s-h a),») SP sein Feasible? Value A (0,%) Tnfes. 6750 S400 ao ee 5625 6300 o Initial PE M[(0,0)]—> Mot cptimal since Cand I [Ce and C1,0)] are better, Conrider te two edges that emanate from (0,0) to Cie) and J tle) Sine the objective lat Z. increnaes ak the same tele (-C? He 5 * M-J: $52"), either cam be chosen, We chrese C. : D 1K,1] js adjacent and better, Move to D, C is not optimal since (Mere that 8 is CP soln but act feasible ) D is not optimal since G L(%,5)] ts adjacent and better, Moreh G, (Note that & is the only better adjacent CPF solh , D is CPF, but werse.) G is optimal since CPF's 0 and T give objective values 5625 < 6000, Mote: T# J is chosen at fest step, the path will bee M-J-T-e (with He same arguments as above). 4-4 Yu-S. a,b» CP soln Feasible? Value 4) Takes S ,%) Fees 5S ox aitral Pt F jnet optimal sime Band D ace better than F. Sina He objective value Z 5 fj inweates faster along tke e; than along the eye F-D (U/q = a we fs Geel 3 ed B is net optim i 7 i s primal since C js adjacent and better. Move te C, CMete A is adjacent, but net Peasible) is eptimal sine B wat Dare ealy adjaceat ¢ PF sel, and C is better than either ” Cc Comer Point __| Profit = 2x, +3 Next St —Comerpoint _| Proft= paths | pear soy an Check (2.50) and (1). 72.5,0) 3 Move to (2.5,0) (i) 3 Check (3.333,3.333). 3333.55) 16.667 Move to (3.333,3.333) Check (3,4). GA 1S Move t0 3,4) Check (0.6.2.8). W628) xg Stop, (3,4) is optimal. Comer Point_| Cost = 5x, + 7x, Next Stey 2.6) 102 Check (21,0) 7 (0,18). (21,0) 105 Stop, (12,6) is optimal. (0,18) 126 °P Ie 4-8 VTrue Use optimality test — here, "better! means smaller, To see this, note Tat min 2 = - max(-2) b) False? there can be an inkiaste number of optimal selations, CPF solos are vot tre vaty pessible optimal solatioas [eg, Mmand-%,t Ky St, Ker, Sle Ky 20, Kaeo Zte10 XT ek xe slo~k — K renying from 0 49 10 are all optimal selativas. ) More that if there ace in fact, more than one vptimal solation thea there mast be aa lattalpe number. ©) Tret: However con't change “may ~ te “twst” because it's possible te have on ue baunded Feasible region where an eative (29 with yust one CPF solution is optimal. 1-6 44 a) The problem may not have an optimal sotution. 'b) The optimality test checks whether the current corner point is optimal. The iterative step only moves to a new comer point. ©) The simplex method only chooses the origin asthe initial comer point when it is a feasible point. 4) One of the adjacent points is likely to be better, not necessarily optimal, ©) The simplex method only identifies the rate of improvement, not all the adjacent corner points. Bei. @) augmented form: Max 4500x, + 4S00Xx, St. x — : 7 x tx, $000%, + Yoee x, +X =! ‘ - $s = e 400%, + Soex, ° 5 ‘eee ae ey = boo 6) 1 X20, karo Xs eo, x, CPE Sor, BE Sols a aaa : ‘oly Men basic Var's Basie Va," , (0,1)1,0, Woe, eo Ny Ge Xa, ae ¢ 2X3, Xe Xe : aa ee so 0) Xa Xe X 1, % 2%, X- oe OX Xs Xe a kaa ks He ea 3 a 4&0, 25) Xa, Xs Nauka Xa Xe : (1,00, 1,100,200) 2, X Ma Xe Xe Fo (0,0) (0,6, 1,1, C000, bow? x 7 a pe 2 kh 3, Ma Ka Xe 4-F Salve %F! > > * x22! ( Hooker, Looe [ Xe =b0e “SOG= 106 _ SOoK,Hx, HLoe } _ Y= Xe Ee Kirks el *) 400 = £06 -See 100 wee {sXe Sooe KF MCER, HXg sare 2 ook, + F00%,2000. | C Kye hs = = tka! Sove ky Fooe%, = 600 SR HYX, HG 400X. t Seek, = dou f 4X r She 0 Aye Ks HE x ) Mat Kat eae) $0bt X,+ 02% © Cove (ee i 2 love -Hooe = 2006 Hoce X47 bee ~Se0e > eee ; 2 MON $00 Ky + Xeeheo) Ke> bee~Hee sold)? 7S wis yy E Sete x, srg ecooe | YOOX +X, 2h00 } FE X.eX2, 26 X32, ket Xg Flooe Xe Fhow 4-8 Xs £200 -$ooe > (ee Xp 4 Lov - Yoo = 206 Gite a) augmented form! Max Xir UX st KB MZ 7 Rieieee oka KPo, Kp %e, Ky Fe, My z 4 by CPF soln BF sala Moa-basie Var's Basic Vae's A (0,0. 0,0,%,4) XX Ma Xy B (0,5) (0, %5,0,%) XX Xa,Xq ie (2,2) (2,2, 6,¢) x5, Xy Ne Xe oO (4ed C4,0,4,0% a Xy YXy BE Sola St Merchasic Solve: 24 A) CP Tafeas Sei Basia Taleas Sel Menbasic Fase vagy E (0,4) (0,4,-4,0) ‘1, Xq XX; F CB,0) (3,0, 0-4) Xr, %3, Xa Xa 5 Magmic E 3% 2tKy28 J KyeB-M eH y2e4 43-1. — Input! 4500 x1 + 4500 x2 subject to » Gis ic 2 Ot ee ea 3) 5000 x1 + 4000 x2 < 6000 a 400 x1 + 500 x2 < 600 and X12 0, X2 20, Solve Interactively by the Simplex Method: 0) 724500 x4~4500 x24 1) 2) 0X3 ° 5000 Xf+4000 xg+ 0 400 Xi 500 X2+ 0 X3+ 0 Xe+ 0 XBH 3) X12 0, X22 0, X32 0, Xa 20, X52 0, Xp 20. 4) 0) 2 0x; + OX5t 0 Xe = 4500 a) a x; 8X58 Xe = 1 2) 0 x: J + 0 Xs+ 0 Xe 3) 4) 0 {+ 500 -%4- 400 X3+ 0 xqr 0 xg+ 1 xq = 200 H1 20, X22 0, X32 0, Xq 20, X52 0, X62 0. 0) z Ox 0x, 0 Xg+1.12 X5+ 0 Xe = 5625 a) 1x+ 0x: 0X4 0 Xs+ 0 XG = 2 2) 0 Xq+ 0 4+1.25 xd+ 1 Xg-2e-4 X5+ 0 XG = 0.75 3) O Xy+ 1 x4-2.25 x. 0 Xq = 0.25 @ (Coxe or (0 K4-0.12 X5+ X12 0, Xz 20, X32 0, Xe 20, Xs > 0, X20 0) 2+ 0 Xy+ 0 Xge 0 X3+ 0 Kg 0.5 X5+ 5 XG = 6000 1) LXy+ 0 K+ 0 X3+ 0 Xq+6e-4 X5-de-3 XG = 0.66667 2) OMi+ 0 Xg+ 0 X34 1 Kgtde-d X5-Ge-3 Xe = 0.33333 3) OK1+ 12+ 0 X3+ 0 Xq-de-d Xe+6e-3 Xe = 0.66667 4) OXi+ 0 Xg+ 1 X3+ 0 K4-6e-4 Xgede-3 XG = 0.33333 X12 0, Xp 20, Xp 20, Ky 20, Xs 20, X20 Solution: (Xi, Xat)= (F, 4) , Bt b000 4-10 43-2 a) Max ®2%, + 2x, FE BK FX, 7 we we +%y = Ki 20, Ee xX,20, Xe zo fe0 —.—rrC~—i“—~™CS—C™CSCSCSCSCSCS Tnerease Xz Crate of imprivement = 2) Lieatering basic variable ler ication |" WO, se X32 F-3K, FO BD X, £% X244 Thas, ky Can be increased te Ye => Xs =0 “Jeaving bari variable Be Meany 1 Using Gaussian Chimination , we get! Z= sXe -3yye “ + Rit 22 TAI +X. +4 x5 2a 3 TER tM FF CK, %Xy 20) Again (0, %5,0,%) is nat optimal since rate of inpctrentnt eX. is So nce : crease Ky, (Xy=0) Xa= F-4%, 20 S x, 8 er —————S—C ——LUrUrSSCS—‘“‘—™sOSOSC_TsSéSsSs 4a 43-v4) (eoat") Again, using Gaussian elimination Now, we are eptimal since xr Z2-d¥,-4%V + |G x. FX, eb “2X3 + $%q 2 *y decrease 2, E “ =€2,2,0,0), 2 b) Computer oatpat! Solve interactively by the Simplex Method: Zz 1-2 m+ 0 xs 0 xy = 0 Ca xf 3 kp xa 0 xa = 8d cy fae = x34 oa 4 = o 1) 2) x o ay 2) XL 0) a) 2) x 20, X22 0, x32 0, xg20. 220.33 xy+ Xqy FQ On aK XY Bo) jncressiag Xx 0 Xq weceld 6 Ge X%2,x322) 0 x: 0 xq = 5. [9.333 Xi 1 X3+0.33 xq+ 0 xq = 2.66667 fo: 667 xs —0-xg-0- 33 Ket = 133333) 20, X22 0, X20, X20. Ze 0 X1+ 0 Xp+ 0.5 X34 0.5 xg OX1+ 1 XD 0.5 x3- 0.5 xg LX1+ 0 x2 0.5 3+ 1.5 xq 20, X22 0, X32 0, X20. KFaR ter, Bed © Agan, you get the same solatien ; optimal solution Value of the Objective Function: z = 6 wiable x1 Xa us 2 2 Sensitivity analysis Objective Function Coefficient Allowable Range To Stay Optimal 412 43-3 Bas| eq] Coefficient of — Right var|wol 2] aX a neeinnen eetaees x58] Nias) 2 ©) xe xs| Bas| Right Yar! Side 2 60 3 10 XS 20] Bas|Eq| _costficient of Right var|wo| 2] 4g %q__ka_ __¥5 | Side @M zfoa} 3 0 0 2 al x3) 1] 0/1.2333 ° 1 0.6667 -0.333 | 6.66667 xl al ol © -2 a oa i} 10 od rye . On xt xp) 2 C0, le, 68); optimal with 24: Fo 43-4 (ourPuT) Solve Interactively by the Simplex Metho« 0 1) 2) 3) Oxgr 1 X5+ 0 X6 Oxgt 0 Xg+ 1 XG 0 a) 2) 3) X12 0, X2 20, X32 0, Xe 20, 520, Xp BO. 0) zet.53 xy+ 0 Xze 0 X540.74 X410.32 Agr 0 XG = 9.09474 $) *tgta ale 0 xgs 2 xgr0l21 420008 Her 0 Xg = 1.68421 CC eCMULDc rLC SPE UG 3) Qiong kis 0 xan 0 xg0.d2 Xgro.l Her 1 Xe = 3.63158 a rrrr—C—S=S=SC= y _ : Optimal solutien ist (0, a, My) =O6 2,5) , 2° 98 4-/3 4369) The simplest adaptation is te farce Xz end Xs inte the basis Centering basic variable) at ‘the earliest eerertunihy, (Zn fact, since we know Gaussian elminahen, Wwe Can Simpiy Solve far the optimal Sola tien Airectty y by 2=Sx,4 3X2 + 4K, 2%, +X. RXR XY 3K + Ky +2%3 tke O% ke, Xp Ha Xs ze) Cheece i+ ints basis but Ht doesn't really matter...) wn zo Be @) Tnerease Xa — we could de Xx First, (%4=%370) Xy220-%2 FO =X, F220 —S my Xp 230-7 Xe FO =D xX. 230 Therease X2 # 2°, Xy 0 Z=-% + Xz -3%y +00 AMF XL KS KY =20 x, PR Ky Xs 210 (41, Xa, X3, Xe hs FO) what we would (i) Taccease Xs (us it turns out, this Is normally do anyway ) C2 %y FO) X2220-X%3 BO =D Xz SZ X52 10 -X3 26 => Xz S70 Se Increase %y te 10 =D Xs— 0 Sin ZrrHWs— Wy Xs t+ FO X, +X + 2%, -Xs =/0 % +X%y hy Xs TO C4, Ky Hs 20) This is, indeed, optima), GI XE xy) = (0,10, 2) | 2*= FO Aly 43-6 (a) Because 2,80 m the ophinal soluhén , we need only sole He lear program: Maximize 2+ 2x, + Ixy Gabject 40% K, t2x, £30 Rt xy #24 3x, +x, 22 K, 20, x5 BO, or, epuivalontly , Mopinige 22 2x,+ Bxy Fubjet tek, + ey £30 Ait Xq $20 XO xy 72. Swe ¥,20 and 4570 ih te ophimal Schchin, they emust be basic sartables th the ephinal colubin . Cheasring % ard % a5 the first two ery Va nialdes will dead directly to the Saluhen, Leaving) variables muse sfell be dele, by a minimum ratio tee. 43-7 A) False. The singlex meted 's on|éa] ——_coetfictent ot | ishe (b> varie} 2} xt 25 | tee = — a I o ole (0) my) + 2 41 oj 30 | %selEa] ——_oufficient of | aipnt Vor|wol 2] x1 RR oX¥ XS | alae —lll | Lal alos | 05 0 15 Of Wate tos ols zo) 0 9 0s Fs sa[ba] —_ortficient of | aight varite] 2] x1 MAY XS | loo ee rt 1 Zlot] 9 8 4 4 [so ase a tt att Te a Gi Xe,%) + 12, 0, 10) fs optimal with 2=5e rale fer choosing Pe entering base variable is used because it Will inerease Z at He fastest cate, because It has fre best rate of improvement, b) Trae, Simplex metad's varighle detemiaes wh Fiegt as He ertering cAeosing a4 © rate for choosing Me fenwing base. Jah basce Variable drops te 2ere bese variable is incveased. Thus, er would normally cause This variabk tb become negative Cinteasible ©) False. qhen Ke simplex mettsd salves fap the nexd GF sola tion, elementary alge beaic operations are wsed te tliminate each basic variable fim all but ene egoation Lits eguatien) and fe Jive it a coekbicient of +! in Hat one egvatica ale HAL, Computer extpat (Tabular Fem) Solve Interactively by the Simplex Method: Bas| Eq cet ficient of Right Var| Nola] % x3 ig %e| side z | o| 1} -4500 4500 0 © o olo x3] 2] 0 1 o 1 o 0 fon} eames) (D) xa) 2] | 0 7 0 1 0 of, t xs] 3] of sooo} 4000 ° ° 1 0 | 6000 xgl 41 oll_4oo} 500 ° ° ° 1 | 600 Bas| E¢| — coefficient of Right Var|No| a] xy __&: xq ig Xe | Side z} ola 0-450 4500 ° o 0 | 4500 xi] a} 0 a 0 2 ° ° ofa (OD xa} 2} a 0 1 0 1 9 cnet: x5| 3] of ——o| ano. 30009 3 000] xgl al ol ——o[_ S00} 400 0 ° 1-200 Bas| Ea| roetficient of Right Yar|wof 2) qe g_| side z] of a ° 0-125, o 1.425 0 | 5625 @ Xi} 1} | 1 ° z ° ° ofa x4] 2] 0 ° o| 1.25 1 20-4 0 | 0.75 x2] 3] 0 o a} 1.25 o_o. 0002 o | 0.25 xl a} olf —o ol 225 o-0.125, abs Bas| Eq Coefficient of Right Var|wol 2] Ximena x EET %q_| side G) azfoi} o 0 0 © os 8] co0o xi] 2] 0 1 ° ° © 0.0006 -0.004 | 0.66667 x4] 2] 0 ° ° ° 1 0.0004 -0.006 | 0.33333 x2] 3] 0 ° a ° 0 “4e-4 0.6056 | 0.66667 x3] alo ° 0 a 0 -6e-4 0.0044 | 0.33333 Selutiext xt Xi7% 2s dove 44-2, Bas| Eq _Coctticient of Right var|nof 2] xX, % a_| Side co) zfofap ae 0 o}o | 1] of 3 z of se] xal 21 0 T x ° Ta Bas| Eq Coefficient of Right ) eas % a] side Ce o} 1}-0.333 0 0.6667 o | 5.33333 Xo] 1] offo.3333] 1 0.3333 o | 2166667 xal 2] olfozesez 00.335, Ti 3333 4-16 44-2, (cont') bas|Eq Right var|No Side @ alo s x2) 2 2 xl 2 2 Solution! me, 44-3. ay 60+ 40 0 15 20 25 30 Optimal value of 2: 60 Optimal solution: ( 20, 20) 4-17 HD cy 5 142 Tee di xy XaF © =D x, =40, xy 7/00 (Hy Hy are slack ee a constraints respectively) Trerease X) Xa sete 0, %3F40-K, FO =D xX, S40 Yay 200 -4%, 20 =D ¥, E25 Eoin Can increase X, HAS D> Xy FO Met, 22 2%, - 4x4 150 2x, t%,-g He 2 15 ve apky 225 Cu ha es, Xy zo) The 2! (25,0, 15,0) is not eptimal, 50 increase Xz, (Xy =e) *37IS-2x, Fo = xX. 420 «min % 25 ~dx2 20 => YX, 2y00 Tnerease X2% 20, X; (leaving basic varrable) > © So now , 2 = ~3%5-4 xy tho Xa t $xy- $A, = 20 SLL Cua, G, 4 22) (2,2,0,0) is optimal , 22 60 4-18 HMB) Solve Interactively by the Simplex Method: peloe oe 0) fe ee ee 3 ee ese i _ 1 20, X22 0, X32 0, Xy 20. 0 Ww a Xz 20, X22 0, X32 0, Xe 20. 0) z+ 0 Xg+ 0 Xg40.67 X3+0.33 Xq = 60 (D4 O Xa+ 1 Xpe1-33 x3-0. 2) 1 K+ 0 Xg-0.33 X3+0.33 Xq = 20 1 20, X22 0, X32 0, X20. 2),F) Solve Interactively by the Simplex Method: Solution, XX = 20,x1220 Nate CoeFGicrents for Xi and Xa are %Q S20 Tnerease Xz fe 1S , X3 gees to O Lleaving basic variable) Zaks “PX, +45 EM TR, + EKs 215 4x, -4%3 +Xy 2S (1X2, X53 xy 20d Ter 2! (0, 15,0,5) is net optimal , Inererse X1, (X20) %,FIS- GX, BO DX, 230 Kee SLX, FO DH, SO TO Trerease KX, te 10, Xy —PO- Ze ry ky + FO Ma xg ky 210 x, Ms t2Ky =O Ch, Ha, hy BO? This is optimal ¢ (xt, x2) 2 Lle,10), 2*= 50 d) Solve Interactively by the Simplex Method: 0) 22_2 Xj 3 Xe 0 xe xy = 0 (2 2) Txge i xgr 03+ 1 Kg = 20 1 20, X22 0, X32 0, X42 0. 0) 20.5 Xy+ 0 Xp+ 1.5 H+ 0 Xe = 45, O) 1) Fosse 3 xp 0.5 wae 0 Xq = 15 2) 0.5 Xi 0 X2- 0.5 Xye 1 Xg = 5] M1 20, X22 0, X32 0, xg 20. 42) _ 0) % Om+ Ox axe 1x @ oy ox: ix int is 2 1x Om ix 2af 20, 220, 420, x42 0, Solation: xJrho, katsvo , Z*= So ©),) Solve Interactively by the Simplex Method: ABZ = 50 = 10 = 10 Ceebttcients <0 > not optimal. Xz i entering baste variable Pivet col = Xa west xy 2 Beisemia Se Xs ts pivet row, Gaussian elim. Yer Coekticient <0 9 not optim, X, is pivot col. Ratic test: x21 b= 30 yt E200 emia 30 Xu is pivet row. AU coetRicients 20, 50 option! Btes0, Xt20, x,42/0 0) z+ 1 xt 0 x3 2) 0.75 xy 1X, 2) Gea od o-xd 3) 71.25 K+ X12 0, X2 20, X32 0, X42 0, X52 0, KHz 0. 0) 241.83 X1+ 0 Xq+ 0 340.83 Xq40.67 X5+ 0 XG = 76.6667 1) 0.333 Xj+ 1 Xgt 0 340.33 Xq-0.33 X5+ 0 Xp = 6.66667 (2) -2)--0.833 i+ 0 Kae 1 XZ-0.17 X440.67 X5+ 0-H = 16.6667 3) =1.67 Xj+ 0 Xg+ 0 -X3-0.67 X4-0.33 X5+ 1 Xg = 26.6667 Xi 2 0, Xo 20, X32 0, Kez 0, X52 0, Xp 20. Optimal solation : (%7,X4, x} )2 (0, 64.165), 24° HF > ere re var|no[ @] my pag | Side o) «= {ofap 2 -« 3 9 oo _ofo Ca] a| 0 fees eee emeen g eee oe eee =| eee] xs| 2] of 2] 2] 2 SOSOS:~SO*SC*SCS CS ral) aj] 6 fb OG AG bas|2q Right var| nol 2] ike | side z| of a oo} 60 WM x] a} 9} 0.5] 0.25 2 o | as x5| 2] o| Pee] xel 31 ol : Opeeeeeialacs bas|=q Right var| Nol 2] ia | side (2%) 9] tf-e223 oo 0.8333 0.6667 g | 76, 5567 Xa] 1] ofo.a333. «= 200.3333 -0.333 0 | 6.66667 x3] 2] ofo.es33 «= 0S 1 -0.167 0.6667 0 | 26.6667 xel al ol-i.ce? 0 «= 0 -0.667 -0.333 -2.'| 26.6667 Optimal Solution: X= 0, Ayh= 6% ‘ , 3 4-23 z 7 z e067 F187 o w @ 0) 2 2) 3) 4) X12 0, Xz 20, X32 0, X42 0, X52 0, X62 0, x7 20. 0) 240.67 i+ 0 X24 0 at 0 441.33 542.33 Xse 0X7 3) 1,333 H+ 0 x2+ 0-341 XG-0133 XB-0.33 Ker 0 XY 2) 0.333 Xi 1X2 0 34 0 440.67 X5-0.33 Xe 0 X) 3) 0.333 Xi+ 0 Xa 1 X3e 0 X4-0133 X540.67 Xge 0X7 4) H+ 0 X2t 0 X34 0 X4-0.33 X5-0.33 Xg+ 1X7 14.6667 3.33333 1.33333 1.33333 0.33333, 1 2 0, X22 0, X3 20, Xe 20, X52 0, X62 0, x7 20. Optimal Solution : XU XE,%5 d= (0,4, Se ea ids Bas| ert i Right var|wol 23a x Hq Xa MSG | Side z| of 1/0.6667 ° ° 0 1.3333 2.3333 o | 14.6667 Xa} 2) 0/2.3333 0 0 1 -0.333 -0.333 o | 1.33333 x2] 2) 0]0.3333 1 ° 0 0.6667 -0.333 o | 1.33333 x3] 3] 0]0.3333 0 1 0 -0.333 0.6667 o | 1.33333 x71 al of0.3333 0 ° 0 -0.333 -0.333 a | 0.33333 . $38 2 3, 2-143 = 747] 13 eselotion 4-25 Right Side (» REET )= (us,0.5, 6) 3 OX RX Optimal: Z€ 22.5 4-26 44-8 Bas|Eq| cosfticient of Right var|Nol 2] jaya Ke Side zJola] 1 --2 0 o ofo co? -wfafo} ap | a as| 2] ofp a a ee) xel 3] ol 2 3L__1] 001 T 50 Bas| Ea coctticient of Right vaz|nol a] =; aa Xa Kg gO Side z}ofajp_-1 3 o o 1 of 6 (DP xa a} o 2] 4 09 «862 (05) 50 x3} 2) of -1] 2a o's | 30 xel ao} fei ecieee Orca eet: meen te|a70 Bas|Eq| Sootticient of Right ga ig [ieee ees geese eimemerye| Pst (2) 2] ofa] 0 3.3333 «=o 0 0.8333 0.3333 | 66.6667 xf ifol 0 eas | 50 POO) Ore | Oona || ce xil al ol = 10.3333, 0S 0.167 0.3333 | 6. 66667 Optimal Sela: XP 6S yoo, RYE MF os 2-68 45-1, a) True, The Ratio test tells us how far we can increase Me entering basic variable bere one of the current basic variables gees fe zero, TF twe variables “tie” for which should be the leaving variable, then both variables go te Ze at He same value of the catering basic variable, Since ealy ge variable can beceme ner-basie in any iteration, the ether will be basic und 8070, b) False, Tf there is ne leaving basic variable hen the selution iS unbounded , (Can increase the entering basic variable indefinitely) eV Teue usually, TF we pivot on the cilumn with the ze coefficreat, then we will get another optimal selutien (remember that The ebjective value will aot change since the "rate of improvement” is Bere, “he exception occurs uhen the problem is ake degsaae H the basic variable (which has a ceclbicient of Bee) is itseHh Cgval te Reco Cie, degenerate), hen fivatting 09 that column wil cot change the aptimal solution, a) False, Examples; (i) Max ¥,-%2 @ max -%1 Woke! Xk. B-1 Kp BO Mee X%,20,K220 Clearty, 2421 Laptinal) bat This Is q trivial: xPeael KEK KELO,) Y-2F 45-2 b) Yes, Optimal Solution: (x, x, ) = (0,10) and Profit = 10. ©) No. The objective function value is maximized by sliding the objective function line to the right. This can be done forever, so there is no optimal solution. 4) No, solutions exist that will make the profit arbitrarily large. This usually ‘occurs when a constraint is left out of the model. ec) gat the obyetive Functiom be 2= Xi- Xa wene initial tableaw iS! (“@ehident at Right Side ae a3? Be | 30 “2 2 7 ti pivet comn has all elements negative Tavs & tS unbounded: 4-28 F) Contribution Per Unit of Each Activity Constraint] Activity 1 Activity 2 Unit Prot ‘Solution The Solver message was that the Set Cell values do not converge. There is no optimal solution, because a better solution can always be found. 45-3 a) 'b) No. The objective function value is maximized by sliding the objective function line up. This can be done forever, so there is no optimal solution. 14 16 18 20 ay 4-29 ¢) Yes. Optimal Solution: (x, ¥,) = (10,0) and Profit ay 4) No, solutions exist that will make Z arbitrarily large. This usually occurs when a constraint is left out of the model. at ex-xiexe Coe Ficient oF Z Pm xa *3 %| Right Side] ra o 27 68 ale toblesu isi a ° ° tsret coluun has all elements negative So 2 is un boonded. Contribution Per Unit of Each Activity Resource 2 Totals Available The Solver message was that the Set Cell values do not converge. There is no optimal solution, because a better solution can always be found. 45-4. Bas| Eq Costiicient of Right Var|Nof2[ a Xj Xa Xx | side 2] ola <1 3 nt o 0 olo 0) | 2] 9 =z 4 3 a © of 20_] ( x6} 2] | 6 se 0 1 0] 40 x71 3] 0 3 3 8 ° 0 aso Bas | Eg Coefficient of Right var| no[ 2] % xy Xe %e__7 | side z] ol a Ble 7 a 5 ° © | 100 CO ml alo =2 ‘ 3 1 0 0 | 20 x6) 2] | 2| a1 3 4 1 o | 120 x1 3] 0 a, 5 aeerear=2| 7 +110] Right 2] eee xj %: xameceon %q x7 | side iv) 4 ° Qnemesge = 33) = 17] o aa | 20 ° a of <6 7-3 © 2 | 40 o o ofan | az 2 z 2 [a0] ° 0 tL 232 ° 1110 Bas| Eq Coctticient of Right var| Nola] x: zane gp | side @) zfs ° ° 074.455 -17 3.4545 17.909 | 693.636 x} a} o| 1 ° 013.545 3 015455 3.0909 | 116.364 x3] 2] 0 0 ° 1 1.0909 0 0.0909 0.1818 | 12.7273 x2l 3] ol ° 1 07.4545 — -2 0.4545 1.9091 | 73.6368 We cansee from eter the seord or the third ifeatlen hat, becauseal/ of te coastratat cock lareats of Xs are aea- positive, it cam increase wihicut Breing abasic variable % 2e70. Prom koa Pore, (hy ,% 1, Ks Ka) = (Ue. 44 38, 73, 44426, 12,930) i3 Feasible Br all OF0 and 69.64/76 is unbounded, AS-s. 4) The constracats of any LP pooblem can be etprtssed in matrmacltion one Axn=b , x20 Te XK, x” are feassble Selattens and xz Baax® us Zoe thea ay ACE, xf)= Eo Axe = Z So X ts also a feasible solution, b) Stace basic Feasible solaHeas are Reusibie solatens (He argument in partla) shows any weighted average of them js alse Feasible. 43d 4S-6.a) TP 2° is the value of the cbjechie fanchén for an ophmal solukin, and x'x%,.., x" a the set of ophoral basis, faéible Soluhsns , Hen for X= za wok fact! 1 zo foe ke P2, N, Problem 44/3 shows x te feasible The objective Fiunchon is of the form “e'W 90 for the feasible sau tin x we have one er(Zaux') pact » Sat ithe: fo x UF alee an ane Soluhin Asa a) x $6 483 1, +3x <6 b) Unit Unit Profit Profit Product 1 Product 2 3 Objective Function (6) Let x be a feasible soluhon which is not a weighted average of te set of ophinal basic feasible Seluhans , x' x?) X% eee ee es average F basic feasible Solutions, no all oP which are optimal. af KX7, Kare Me basic Hesible soluhons which are nat at we can express x as! with imal , xe Zax «Bas eo ay 20 for kL... fer ie A620 for EF 1,2,.,b and someh; #O, de can conclude , 1 ox oe xe Be zi) so x is nd an ophmal soluhén, Multiple Optimal Solutions fine segment between (0,2) line segment between (3,3) & line segment between (6,3) & (6, Tine segment between (0,0) & (6,0) line segment between (0,0) & (0,2) G3) (6,3) 4-32 ) Objective Function: Profit = -x, + 2x, Optimal Solution: (x, a) = (0,2) and Profit = $4 Corner Point (1x, .%)) Profit = -x, + 2x, Oo 30 (0,2) $4 (3,3) $3 (6,3) 30 (6,0) $6 d) The initrel tableay és: = [x [x | %s | % [*s Tight Side ol el-e[ofe fee oO ‘ ° ‘ ° ° 6 ° ° t ° t ° 3 tf" BI) eo} of. 6 a LCrhLUr o}|-alo}lsejo]e 6 eo | Bflolo]e “ts 4 ° ie 1fofe % 2 SS He onigue optimal seluhac is (oz) 4-33 Coceients (row t) for X,,8iX3 art 0, we can pivot to get ot Right side @) Tableaux (2) -(5) Show the set of ophina! solubons: (Ce, X4, 5,%)) # (0,3,0, 2) (2) 04) %,%5,%9)= (3,0, 2,0) (9) C5, 85,49) + (3.0,0, 0 ® le ryx5,4)* (0,3, 2,0) a with B25, 4-34 “eae ) nas|zq| coofticient of Right Var|No| 2! x1 x; x3 Xa | Side “iyi (2) s}ola =2. =3 2 o | -3 x3 1] 0 i 2 1 ofa —) xa! 2i ol 1 7 0 rT The inlite hasic feasible solution is Ckuk.X%y)? (22,43), © Bas coofticient of Right Var a a | side rE “iM z 0) eetsee 20i| a Og aos oo | 3 xa ce a Bas|Eaq| Coefficient of Right var|nol2] xj x23 Yq | Side Tit Bafdy 0 o 2 Br mfafol oo aa a mi2lol 2 0 + a2 yt) es oe Optimal: OM XE) 2 OV, B= F 4-35 4.6-2.€)b) sas} al Coefficient of Right Var|wol 2] XX %e_|_side Tio _~4 S70 (0 gz] ofaf a -2 3 os ° 0 | -600m ¥5] 1] 0 2 3 4 2 1 o | 300 Xel 21 0 = i i = 7 T1300] Tnittal artificial 6F sel is (0,0,0,0, 308, 300) Bas| Eq Coefficient of Right Var|No| a] _% X: 4 __¥ ¥g_| Side =2-75M-3. 75M -0. 7500 Toasm | ~225i (O zfolal oo “us “2's Vi2's oo “sos | vis0 ¥s} 1) off 2.757 3.757 0.75 1 =0.25 [225 xl 21 0 10-125 [0.125] 0.625 00.128 137.5 Bas| eq] ficient of Right var|wo| 2]; zjeeene x eae Kg_| side imine (OW z] oa © 0.3333, 0-2 40.667 +0.333 | 300 x3] 1] 0 00.7333 if[-0.2 0.2667 -0.067 | 60 xql 2} ol[ 100333 oO] 0-6 0.033 0.1333 130] as | geen Coatficlent oteee eel Right Var| Nol 2] xy x: xqneenex eee %__| side iy iit (3) 2] 0} a].3333 o.aaa o 0 +0.556 +0.778 | 400 x3] 1) 0~0.333 0.7222 1 0 0.2778 -0.111 | 50 xal 2l ol1.é667 0:0556 ° 1 -0.056 0.2222 | 50 Optimal! (X1,X4,%5, x5) > 0,6,52, 50) , 272 400 ©) -#) phase 1: Right Side @ 600 300 [300] Fritsal artificial BF Sol» is (0,0,0,0, 320,300) Bas| Eq coefficient of Right Var|wol 2] Peer %g_| side (2 ] ofa 0 -2.75 -3.75_-0.75 o 1.25 | -225 Xs] 2] 0 oO 2.75] 3.75] 0.75 10.25 pas] xal al o 10.125 [0.125] 0.625 00.125 [37.5 Bas| Eq| oef ficient of Right var|No| 2] x: amen Xe] side (D2 1 ° ° ° ° 1 ilo x3] 1] 0 0 0.7333 1 0.2 0.2667 -0.067 | 60 xil 21 0 2 0.0333 0 016 -0.033 0.1333 | 30 4-36 Xa 01.6667 0.0556 oO 1150 ‘) Optimal Salm 2 (X1,%2,x3 xy) = (0,0,50, $0) , 2400 4) The basic seluhons of the tuo methods coincide. They ace arhfietal basic feasible solbions for the revised preblem unhi! both arbPeial varables x and 1 are dewven out & Ha besis, wh In the Two Phase Method js tha end of Phase 4 h) Variables Constraints 464 Variables Constraint 4-37 46-3 a) b) 0) (D @ ©) Maniniv2e 62) * ~2y,-3y¢-2, Pubject te a “Bate, Sob % 20, x, 20, 70 Bas| eal coefficient of Right Var|Nol 2] xa: x: xq Xs | side =a "626 | (0 ite er o o | -14m Se] 2 of T 7 cet 0 T ots) x7l al o 3 2 ° ot ° 116 Bas| Ea| Right Var| Nola] Xa X_ Xgq___¥7 | side ~2 5H IM 0-54 TS ~2i z | of a} 42225 0 -0.5 40.75 im -0.75 o| -6 Xa] 2] of 0.25 4 5 -0.25 0.25 ole Xa 21 0 5 oO a 0S 112 Bas| Ea| Costficient of Right Var| ol 2] xq xy ¥g___%7 | Side init z] of a ° ° 0 0.5 05 -0.5 -0.5 X2| 1] 0 0 1 0.6 0.3 04 0.3 -011 | 1.8 xl 21 0 a 0-04 0.2 014 -012 0.4 | 0-8 Optimal | ONES) (0.8, 18,0) , 242% (fFiveting Xs Inh he basis Ge Xa paviaes the aHernake eptimal solutton (2,0,3) , Phase asl Bq Coefficient of Ri ight var|No[-@] 4 X45 Xe | Side o) Zz) ofa} 4 6 1 a 0 o | -14 (Se) a) of ap 9 tf xl alo > 2 ° or ° tte Bas|Eq|_ coef fic: Right ver|wol [gpg ge eS alo ° 1 -0.5 1 4s o| -2 2} 4 10.5 -0.35 00.25 ol sol 2 os at fei 4-38 Wb-3. ©) Ceoat') Bas| Eq Cosfficient of Right Var|Nol @[ agp ge | Sito (BW zfolal 0 0 0 0 o 2 ifo X2] 1] 9 ° 1 0.6 -0.3 0.2 0.3 -0.1| 1.8 xil 2 o! 1 0-04 0.2 -0.6 -0.2 0.4} 0.8 Phase 2: Right side 7 Le ole Again, optimal is On¥, xf n})2(0.8,1.8,0), 2%=7 with He alkenate optimal (by pivoting Xy inte He basis for Xz) of (2,0,3), @) The basic selahons of the twe metheds eomeide. They are arhferal basic feasible Selahins fer the revised problem Unt] beth artificial variables X% and xg are dewen out of He bass, which m phe Taso Phase Meted 's the end of Phase 4 4 . 7 ]Optimal value Variables Constraints dad. HE. once al! anhficial variables are driven from the basis in & matiminaion (rinininabin) prolam, choosing an arhifiail variable to re-enter He basis can only Inver (raise) the objective value by an arbitrarily lange amount deperding on M 46-5 2) Contribution Per Unit of Each Activity Acti Unit Prof Solution 4 Solver could not find a feasible solution. 4-40 Bas |Eq Coefficient of Right VE varjtolic) a xi) | x2pixape tsa 6 side 2} 0} 1] -90 -70 1.0e6 0 1.006 0 \) xa] a} 0 ea 1 0 0 2 x1] 2] 0 1-1 o -1 1 2 Bas Coefficient of Right var|e| 2 X20 X34 XS side 2 0 -25 1.006 0 1.006 90 x1 1 0.5 (0.5 0 o rt x1 0-15 -0.5 -1 is 1 Bas|Eq Coefficient of Right Var|No| Z| Xl 2X3. X45 side Pete ole 0 1.5e6 1.5e6 1.006 oO -1e6 x1} 1] 0 ft er0e noes 0 0 1 x5| 2] 0 OF ies) 05 1 1 3) bas Eq Coefficient of Right var|No] Z| X1 2X3. 4S side zl ofa 0 ° 5} 0 1 0 p Fol YEG 2* 1 1 0 0 2 x1] 2] 0 1-1 oO -1 a 2 Bas|Eq Coefficient of Right Var{No| Z| Xl K2 x3 x4 XS side zfola o o a ° 1 0 f\ x2] a] 0 1 0.5 0.5 0 0 al x1| 2] 0 Oar oy a} 1* a Bas|Eq Coefficient of Right Var|No} Z| Xl x2 X39 x4 x5 side Benefit Contribution Per Unit of Each Activity Minimum Activity Activity 2 Solver could not find a feasible solution, o} Bas | Eq| Coefficient of Right var|No| z] Xl x2 x3 X4 XS x6 side z | 0| 1] 5000 7000 ° 0 1.006 1.026 ° () #1] 2] of -2 1-1 0 1 0 1 x1] 2] 0 it -2 Ot 0 1 a Bas|Eq Coefficient of Right Wae|No| a |) X11 u2ueetasmexere: x5.) 1x6 side fi) el eo|la 0 17000 0 5000 1.026 1.0e6 | -5000 x1} 1] 0 0-3-2 =2 al 2 3 x1] 2] 0 ae OQ = 0 1 1 Bas|Eq Coefficient of Right var|No| z| Xl x2 K3 X4 XS X6 side (t) Z| 0| 1| -le6 2.006 0 1.0e6 1.006 0 ~1e6 xi] 1] of -2 1-1 0 jaro 1 x6| 2| 0 ed oy 0 1 a 4-42 (eomto) Bas |Eq Coefficient of Right var|No| z] x1 x2 «X30 X4 KB XE side z | 0| 1]1.0e6 1.0e6 1.0e6 1.0e6 0 ° -2e6 Q xs] i} of -2 9 1 7-1 ° 0 1 0 1 x6] 2] 0 1-2 ot 0 1 1 *) bas Eq Coefficient of Right War |Wo] |e xi] | x2eexsiexoe 5 x6 side z | oj 4 0 ° 0 ) 1 1 o fo) Hf 1} of 2 5 ot 0 1 0 1 xi] 2] 0 1s 2 0-1 0 1 1 Bas|Eq Coefficient of Right var|No| z| x1 x2 x3 Kd x5 x6 side Biase ° ° 0 ° 1 1 ° (\ x1] 1) 0 oO -3 -1 -2 is 2 3 x1] 2] 0 2 ee 0 1 1 Bas |Eq Coefficient of Right var|No| Z] Xl x2 X3 X4 X5 x6 side z | ola ° 3 1 2 o = -3 It) xs] a} 0 3 rt 2 1 2 3 x1] 2] 0 1-2 0-1 0 ae 1 Bas |Eq Coefficient of Ri ght Var]No] 2] Xl x2 3X4. XS side 6) z] ofa 1 1 1 1 0 0 -2 ) x5} 1] 0] 2 1 oi 0 1 0 a x6| 2] 0 1-2 0-1 0 a 1 44-7 a) asl al cs of Right Var| Wola] Xi ky %_ | side <3 26208 > 2) oapo-2 5-3 aw o o_| -70m Xs] 1] 0 a7 2 tt z O30 xel al o 2 + T ° ° 1150 Tnittal artificial 6F Seln is (0,0,0,0,20,50) 4-43 - ,) Bas} Ea] Coefficient of Right Abt b> ee eta x x see Xe | side Ta he a =108 (oz) o 8 ta 0 | +40 xi] 2] 0 af—-2] a =i 1 o | 20 Sel 2 It Se =a 110] Bas[zq|________ coefficient of ____| Right var|No[ 3] Ka | Side. 2) 141 @ 2] ofa o 0 -2.125 0.25 -0.25 +1.125 x] a] 9 1 oy -0.78 x2l 21 o 3 T <0.28 0.125 Bas| Eq coatticient of Right var|No| 2] mj ag | Sie Tt 1st (3) 2 | of a}2.8333 ° 0 1.167 41.167 41.833 | 115 x3} 1} 0|123333 ° 1 0-667 o.6667_ 0.3333 | 30 xa] 2) olfpzt687 z oo. i667 0.167 01667 15 Bas| Eq] Coefficient of Right Vaz|Nof a]; ka Ce | Side. TH @ 2] of a 4 7 o 0 m3 | 150 x3] 1} 0 2 ‘ 1 0 ° 1| 50 xal al ol 1 6 0 4 «1 i | 30 Optimal Solution (%,X3,%3) = (0, 0,59) , e4= 150 c) Phase 1: Bas| ) a) ee aed |S Var|nol 2. qj Xe | Side (Dz of a x} a] 9 xel 21 ol 1 pas] eq » {2 x cea ¥E_| side. ¢ Zz] oj. 0 oO ° oO 1 1jo x af o| 1 0 0.75 -0.5 0.5 0.25 | 22.5 x21 21 o| ° 1 -0.125 0.28 -0.25 0.125 | 1.25 Y6-F Of) Phase 2: Bas) Eq) coefficient of Right Var|nol 2] 4X: Xa | side 150 50 30 Optimal! CHE, x5)? (2,050) , 242/50 G) The basi soluhins ff tHe two methods coincide . bee artificial base faasrble seluhons for the revised problem unfl both arb ficial “vacvabks x5 ard % are drwen out f the basis, which in the Two Phase Methed is the end of Phose 1. h) 750] Optimal Value Solution Right Hand Side Constraints or 20H 50 "=" sat 445 Q) b) Variables Conetrants OO} z Se-20 1e-19 ° o 1 1 | 2e-18 0.5714 -0.571 ° 10.7143 -1 | 20 0.7143 0.2857 1 0 0.1429 0 | 60 Bas| Var] (3) 2 x1 a3 Optimal Solution: CXIx8,x3 = C35, 0,35) , Bee 17S Piveting X. int We bass Bre X, provides fhe alternative cptinal solution (40, 35,0) 4-46 8) 175 |Minimum Value Variables Constraints (6-4.a) Right Side oo) ee 125 o Bas} Eq] Right Yar] No| X6 | Side it i 2] ofa} os ° 0 0.5 -0.5 -0.5 | -90 s|f 2) eo] 0876s 2 0p 0c 0%25t 08s) | =ta5q) ic) xal al of -0128 © 10-0175 1:25 0.75 | a5 Optimal is X20, AXIS, APTIS, Z¥= IO o alo 1 2.6667 2 ela 00,3393 Blo ra] (coud) Optimal: (XX4,%5) = (0,155), B%2 70 (©) In both the Big-M method and Te Phase Methed only the final teblag (2) vepresents a feasible Soluhon for the real problem 3) 3 2 « 90 Optimal Value a uv 1% 6 Right Hand Side Consens =—Ft J co aol af aoe aa s Right 46-10 a) Bo sige lo) Tae Tie 43 2 97 om ° o | -20m =i a 0 o 1 o | 10 2.1 iesene=a 2 aio] as | Ba Coefficient of Right Var|Nol-@[ xq pg ye | Side. a) ~O-5N~0.5N 0,548 0.5m | -i5 Dz] ola O 43.5 45.5 41.5 o-1.5 | -15 xs} a} 0 oT 0s] 050-8 10.5 [is xl 2l ol TLo5) 0.5 0-8 ost Ss Optmal : (x1, X4,%5) = (2¢,30,01 , 2% = 120 b) Phase 1: Bas| Eq] Coefficient of Right w OY sto a o -0.5 -0.5 0.5 o 0.5 | -15 Xs] 1) of oT 0.3 0-3 0's a 0-5 | is] Scr 210) pee lee-076 frog: S pas Omen eens) Jas Right side ° 30 20 1-1 0 el WIRE, KF) = (20, 30,0) , H=120 1 ° ° ¢ ©) Only phe Final tableau fr she Big-™M methed and 2riese ime ed cepreseat feasible selufiens to rhe real problem, The BE solutions forthe 2 meteds are the same. 120 fone ght Hand Side a\ Soon tore t0>= Constraints H6-1 a) False, The initial basic solution Sar the artificial medel /s not feasible for the original model, 6) False, The reverse is true: if at least one of Ke athc,a/ variables i+ nat zero, then He real problem has no Basil salatsoas ©) False, The two methods are basically oo pan ae eguivalent — they should take the same # of erate, las WG-12. a) Substitate X7- for % yt Maximize ZzXt- ky +4X, 4 XQ Ys HX HX. +X ES “NP XT 7%, +2% Slo XPzO, XP 2e, HBO, Ky ZO b» Bas| eq var|No z| Xs} 2 xel 2 bese — costticient of Right Ver|Nol 2] mya Xa Re | Sie z | of 2 5 0 2 4 of2 Xa] 1] 0 sq]: to xel 21 0 shea (cour) 4-50 e-(2 (cor?) Bas} Eq] ficient of Right var|wol 2] xq: xy kg x XE | side z| ofa ° ° ° 5 1 3 | 35 x3] 1] 0} 0 ° 1 1.8 0.2 08] 9 x2) 2] ol 2 a fot oro] -ovae 021123) Optimal Solution Cm the original problem): 5 CB XE x) 2 190), 24235 (eit o Rta) a Solution Constraints 4643 a) a tb) TP we let xa, %i%e od Frew 132 Hy , we get Maximiae -X, X_ t4xy - 12 Dabject 6! ~3q 43x. +x, £4 Hy rky t2xy £10 X20 Ky ZO Hy 20. SSS SRI 14,2-59) is optimal with 2* 143 a Bas] Eq| Coefficient of Right var|wol 2] Xa aK | Sige 0 eee eee 08 (0 () xl aol 3 3" xs} 2] of aa a fo a a] Bas|Eq| Coefficient of Right 0 afer mee xr xem gemma eg cp zfola 3 3g a xl 2] (asp agp eos a 3 2150 lpes05 (==0-61lieeent penn One SglnS (cout'D) 4-51 4.613 (conr'y) 23.4286 1.14286 5.57143 0 0.8571 1.5714 1 1 0 0.2857 -0.143, 0 ° 10.1428 0.4286 Optimal for ariginal problem is ° OU KES = Cb, 2.58) ZED (far revised pablem + (0, 114,554) with 2523.43) Ye-1$ a) Lee oro” Xi-%ey Xzwp 7 Xp ky And yoy? Xs-%y Tre equivalent lear preqrtm with nonnegativily constraints is Maximize 24 -n, 42, + 2ay-2%y + Xe -% Subject to 3x, -3xy tap-% S120 Hye “3 + Hy “Yas Hg £ BO “3x, 13a, + Ry Hy *2ay 2x, £100 Ke 20 fee 6242, 6. b) Bas|zql Right Var| Wo rom Hi Kg Xs KEK gg | Side z}ojap oa a2 2 do dt Cod x7) 1 of 0 0 373 a 2 o ‘of 120 xa] 2] of a -a]-1] 1-440 10 | 0 xg} sl ol -3) 3L_ af -1 2 -2 0 o 1] 100 Eq| Right {nol 2] xy xy _Xa__Xq | Side of ap aan 0.667 0 of 80 a} o of—o 01333 0 0 | 40 2) of af =a 0.333 1 io | 120 al ol 3 a BT Bas| Ea| Right var|wol a] aes TLE 2 2/42} 9 0 0.556 00.333 | 100 (2% xs) af o o 9 Jo.333 0 o | 40 xe] 2) of oo lp.222 10.333 | 140 xgl 3 of =a Fo.i1 00.333 | 20 eourd) 4-52 461g leon’d) Bas|Eq| Coefficient of Right var|wo| a] Xp aX Xs Xs Ka Xa | Side. 2] ofa ° 0 0 0 6,0% 0 0.571 0.071 0.357 | 110 (3 xs} a] | 0 ° 1-1 3e-20-3e-2 0.357 0.107 0.036 | 55 x6) 2] | ° o 0 0 -2 10.071 0.321 0.107 ] 45 xal 3} ol -2 1 0 0 0 06 -0.07 0.179 0.393 | 45 srimal for He Optimal fer revised problem is (0,45, 55,0,0,45 ) 5¢ of origina) preblem is: (n%, x2, xp) (45,95, -45) with Bho 110. F40 |Optimal Value } ‘) souion Right Hand Side Constraints HOWE a) sp ise uavsh te decrease the chjechie value in the Singh Meth we will choose as entering variable He nonbasie variable that has the largest posihve cosffiewut “in the objectiv€ row of the tableau. We Will peed Phe Same minimum rabid Hest fo determine He lowing variable , however bb) basieal coettictent of | aint ariel 2] x1 x2 XS XS X67 | sige Sasleal coetfictent of | rine Var|Wol 2] x1 x2 XS Xe HS XT | side ZI 0} 1] 0.64 0 0.0.643-4.216 0.64 1.21 | 130 (2) x5 1] 0}. 0 4-0.36 0.216 0.357 -0.21 | 10 XQ] 2] 010.857" 10 0.145 -0.29 -0.16 0.286 | 10 4-53 Coefficient of | Right M2 OS XH OXT | side i nL 0-0.75 0-075 10.75 1 | 122 00.7% 1-02 0025 0 | 175 TANG? 00.167 -0.33 -0.17 0.383 | 11.67 (x, %a1%3)* CGF, 0, 17.5) 75 ophmal with &* /22 6-16 seated coattictent of | nia (gael a ow | oe Masiniee 22 - 2s, *2aa + oe 1 Pen Crew a al Haye 3% zl 2 2 4 3 0 9 0 ole subject te wi t}o) 1 ot 30201 0 0 Of Wome tty +3ey ede F wal 3} 0] 22 en 01 0) Onl cnen tw otest TF . oe 0 8 0 11 Ax -2x_ +X: 22 Cee ‘ coefficient of | Right Ky + ng F2xy ty tame FR « Nariel 2M KOS KET BAD | Side PO fiat 20 0 0 Of? 3) seston Constaiats H.6-17, ReFormulation ¢ Mayimi ze pth, +2¥%3 — By +z 220 5%, TbK2 WS%d tks = 50 Ky +3, THK +X = 30 xq Xe___¥7 | side 1 ° ° 0 | -20 “1 0 0 1| 20 0 1 0 o | so o 0. 2 oT 30 ee IF ont DD Bas| Var! Eq| Coefficient of Right a fficient o Nol 2] a My Maks GE | Side > Coefficient of Right var|Nol_2] 1 x xy Xa xs Ke ¥ | side z] of 00.5375 0 10.0375 0.4375 =o | 5 Sa] a] op=O°S36 see (oles me1@=080aulmovasn t= eral xi} 2] | 10.8625 0 00.0625 0.0625 0] 5 StaHe3 e018. 0) 0%as75) 21 111 o}080is 08ae75)¢ «104 Since this is the optimal for Phase 1 and the artificial variable X,= 5 0, this problem dees nat have a feasible solution. 4 Tek. CC ie 218 % GO| eo) (a) xX aol 4-SS GAeL (cont 0) me ee (2.0) remains Rasible —VYthe cP Selo (B55, ¥) remains eek BERK, For KSA GL feasible fer 2K eK ESR EIS © shat CP Son beveme in hacible, . : (2,6) 15, ef codese, Me eptinal. payer? 12 \ ed) the cP Sein ($74, rEmans feasible fer Beit 2R22K $k S24 | —> (ateasible fer other valacs 0 RY Re 0 et dHeed eit ate Gashraint {aaa ine) KitBKy BD F=(a)7 VAY z6 yi 29 BDA) Consteaint (Hae) ay a Ze(verearyee 7 * ent hn g25 Be (BrAaye? J Yer ky) rf Me af 4-56 41-2 b) From (a) we see that 6,28) be74 are sensitive parameters, As will be seea in part Ce), C,21,€,7 2 ave net seasstive parameters, As can be seen in the Figure in part a), bath constralats are active ee bending , 50 all coefficients 420%. 7 3, FF I, F227 bo are seasihve parame tes, ce) We see that the eptmal solution remains Me same tar? 420,42 (ey fined at 2) and 5 $6, 43 (c, Gredat 1) yaa, ont’) Locking at He ta" Maes, the CP Sd Pa0ge% Bam (4,9) t (o4) a5 by ranges Som Hw IR. The CF Sen becomes jnfassble fur values of bb outsrde of this range. cits tnoking at he Lr From % t 8. the allewable range sore there fre, 4 8b, 41% and eb, 28 © ]Optimal Value Varabies Constraints ‘Adjustable Cals Final Reduced Objective Allowable Allowable’ Coll__Name Value ___Cost_Coefficient_Increaso_Decrease $es2 OT 88888, a 2 0. 2 1 1 cell ES SESS Ys 4, 7-3 a) Optimal value of 2: 38 optimal solution: c 3) “sg Corner Points T 8, 3) i 8, 0) i 2, 5) ( 0, 5) ( 0, 9) ale ct % Changing eesource 1 te IFunits ("— — a hae) Causes Zt increase t gz y(2)+22)* 39 Se AZ: =1¥ Increasing cesource 2 te 13 units (m= hae) jncreases Ze Z24(8)+ 2(2)=38E. So Az: yr} Since constraint 3 not binding, YA=e. ©) To increase Z by 15, we would need ro increase resource 1 by Fe $ 249 (Solving the LP with resource 1 set te [+9225 Confirms that 2¥= 53), Bas| eq Coefficient of Right a4 a) verlnol-a[ x x Xa x5 Xe | Side z2}oja} 0-2 7-3 o ° ofo (2) xa} a} | 2 (eet 1 0 ola xs} 2} 0} 4-3 0 o 1 ol2 xg! 31 olf 3 2 a 0 0 a3] Optimal | (x4, X45, %4)4 (05,0, 45), BY, b) The shadow prices fer the three resources art given b, The reduced casts (in the objective function) fer tHe correspinding slack variables(circled above), “The shadew prices for resources I, Land 3 are O,25,and 3, rey, these represent He rate at which Z Che objecture value) increases as the corresponding resource is increased , For example , increasing researce 2 by Luart should jncrease EZ by 2,5 Cassuming no other wastraints cause trouble...) Joptma vas ‘Adetable Cote SS Final Reduced Objective Allowable Allowable coll_Neme___Value ‘Cost __Cosfitent increase Decrease ‘Fees Varetles OS gees. 'SCS3_Varables 0. Ss 5 E90 $083. Varabos as 0 3 2 3 ‘$083 Variables gs os eS Contains — Fad Shadow Conaralnt AiowabioAowabio cet Name __Valon_~Price_ RM Sido_incroase_Decrease ete EO ‘SFS6_ 28s 7 1E+30_ 2 S87. 3 3 3 1E+30) a5 = Eo 4.6 co ent of Right ey XE] side z}ofal -2 2 ° ° ofo (ov ¥4] 2] of =a 2 rt o ofa xs] 2] 0 21 2 x ot xXel 31 of 1 1 0 0 1712 Bas| Eq efficient of Right var|wol 2] x %, Xa Xa aXe] Sige of a aeons o o 3 ole @ x4} alo =3 2 0 2 =r oy 2 x3 2| 0 alee T 7 T ope xsl sl of = -5 4 ° . 3 ile Bas| eq Coctticient of Right var|nof 3] 0g) Ky gid z] ofa] 2.5 ° 0 05 25 o|7 (2 x2] a] of -2:5 1 0 08 -0.5 ofa x3| 2] of 0:5, ° 1 0:8 0:5 of 3 xel al o 1 ° 0 2 OF Blea Optimal: OF XE,%3) =(0,1,3) , BL F b) he shadow ree ace *, * Wee, We, yo They ace the macaina "ales of resources 1,2 and 3, respechiely 2 z z z Tlocimal vale vais] 04 ras} Consra TTF se 4 2 44 ao F| i wos 4 Adjustable Cots Reaves Ce Name (Cost _Coeficient_ncresse Decrease asa Vara 2 36 ‘Ses Varwbes—— OED $3053. Variables i 21 Sseb07 1 S083 Varablos 3 3130) i Constants Coll__Name Shadow Constant Allowable Avowabie Price AM. Side _Inerease Decrease 335 Constr a 55: 2 0 2.3333, 0 0.6667 0.6667 1 1 0.9167 0 0.8333 0.0833 0.25 0 0.25 10.5 -0.25 0.25 | 3 Optional ; Ob Xs, Xs Ra) = (0,30) F#2 52 DD tre shadbo prices are hs % ard tel. They ave He marginal values ‘of resources Jortina vue Final Shadow — Constraint Allowable” Aowabie Call_Name_Value_Prcs_RLH: Side Ineresse Deco, Sth tae —_letun___rcw__RuN.Side_Incroase_Decrense, ‘Sass Conevas "_s6668887 2 ete $86. 36 t ‘6 Sse si 4-63 Problem 4,9-1 Linear Programming Model: Number of Decision variable: 2 Number of Functional Constraints: 4 Max Z = 4500 x1 + 4500 x2 subject to D lxd+ OX2< 1 2 OxL+ 1Xx<— 1 3) 5000 x1 + 4000 x2 <= 6000 4) 400 XL + 500 x2 <= 600 and XL >= 0, x2 >= 0. Solve Automatically by the Interior point Algorithm: (x1, x2) = (0.1, 0.2) and Alpha = 0.5 rt. | x1 2 | Zz 2 | 0.0 0.2 |" 1350 4 | ,0;2999 | 9.sg008 | ago9.91 3 | 0.3: 5186.35 4 | 0.4 5503.76 se 5690.12 6 5842.42 z 591-71 9 5971:35, 10 | 5984.91 ni | 5992.4 BY 5996.2 B 5998.1 #| 5999.05 15, 5999.52 Problem 4.9-2 The linear programming problem is: Number of Decision variable: 2 Number of Functional Constraints: 2 Max Z = 1x1+ 2x2 subject to D LX1l+ 3x2 <= 2) LXL+ 1x2 < 4 and XL >= 0, X2 >= 0. Solve Automatically by the Interior point Algorithm: (x1, x2) = (0.1, 0.2) and Alpha = 0.5 It. | xi x2 | z 7 | a1 | 02 | 0.5 1 | 0.24587 | 1.36864 | 2.98988 2 | Q-2s0s1 | 1.97283 | 4-20237 3 | 9:26482 | 2.27423 | 4-813227 4 | 0.28233 | 2.42047 | 5.12328 5 | 0.32398 | 2.48263 | 5.28924 6 | 0.43489 | 2.48368 | 5.40225 7 | 0.82513 | 2.37261 | 5.57036 8 | iv4aze | 2-47597 | 9.77488 9 | 1.72185 | 2.07758 | $:87702 10 | 1:86959 | 2.03012 | 5.92984 HL | 1°34077 | 2-00909 | 595804 12 | 1.97327 | 2.00166 | 5.97659 13 | 1.98735 | 2:00011 | 5:98758 14 | 1.99373 | 2 | 5.99373 15 | 1.99687 | 2 | 3199687 Os Cases hel a) The fixed design and fashion costs are sunk costs and therefore should not be considered when setting the production now in July. Since the velvet shirts have a positive contribution to covering the sunk costs, they should be produced or at least considered for production according to the linear programming model. Had Ted raised these concems before any fixed costs were made, then he would have been correct to advise against designing and producing the shirts. With a contribution of $22 and a demand of 6000 units, maximum expected profit will be only $132,000. This amount will not be enough to cover the $500,000 in fixed costs directly attributable to this product. b) The following insight greatly simplifies the analysis of the problem. The production processes of the various clothing items are not all linked together. We can separate the clothing items according to the materials that are used in their production and instead of one large linear programming problem we can formulate 4 smaller problems. We use the term net contribution of a sales item to describe the difference between its total revenues and variable costs. The net contribution does not reflect any part of the fixed costs The cashmere sweater is the only item consisting of cashmere. The net contribution of, one cashmere sweater equals $450 - $150 — 1.5*$60 = $210. TrendLines can sell at most 4000 sweaters and has 9000 yards of cashmere as raw material. It is optimal to produce 4000 swemers using 6000 yards of cashmere yielding a net conrbuton of 40008210 = 840,000. The silk blouse and camisole are the only items using silk and no other materials are used for these items. We can determine the optimal production amounts of these two items through a simple linear program. The first constraint models the resource limitation in the production process that Katherine has ordered 18,000 yards of silk. The second constraint models the production condition that whenever a silk blouse is produced automatically also a silk camisole is produced. Finally we must include the stated upper bounds on the number of silk items we can sell £66 78000 ~8000 [<= ° 7226000 TrendLines should produce 7000 silk blouses and 15000 silk camisoles yielding a net contribution of $1,226,000. We can determine the optimal production plan for the items made from cotton in a similar fashion, There are no demand limitations for the cotton items. 4-57 Unit profit ‘Solution 3. 4 (Constraint RAG 5 <=] 30000 6 |_| “production = 0 7 & ‘TrendLines should produce 60000 cotton mini-skirts but no cotton sweaters yielding a net contribution of $2,025,000. It remains to develop a linear programming problem for determining the optimal production quantities of the tailored wool slacks, the tailored skirt, the wool blazer, the velvet pants and shirts, and the button-down blouse. We include four constraints for the resource limitations on wool, velvet, rayon, and acetate. Upper and lower bounds are given for many items. When there is no lower bound, we insert 0, when there is no Upper bound, we determine a safe upper bound as a consequence of the resource limitations. = = = z E H elton! escuica Usage Per Uni of Each Renny ‘Aci | esau awe sins [ta ont — wesley iets svt si] 6a uae | Tos | “[avalabe 2 2 @ 2 00 — k= ¢s000 | cr E13 6 % $5000 | Sa000 | ° 2 2 4.5] —s0000 f= | 30000 $ 8 peolanis é S000 20000 Taran as8 las aT pans farrt9a 3 ‘4200 séee.seeee7 ‘sooo 0 Go00 s2adaaaaaal Meinl 4 ama_—| 2000 | 2 31 —IMesimumt — 7000 [20000 —~-go00 | ~Sa00 | good | 20008 a Cs Totals SUMPRODUCT(C7:H7,CTEHT1) _| SUMPRODUCT(CBHE.CTATT) | |= SUMPRODUCT(C9:H9,C11:H11) 70 /=suMPRODUCT(C10:H10,014:411) T ‘TrendLines should produce 4200 wool slacks, 8066.67 skirts, 5000 wool blazers, no velvet pants, 6000 velvet shirts, and 9244.44 button-down blouses. The net contribution of these items equals $2,771,933.33. (Of course, TrendLines cannot produce two-thirds of a skirt, so the actual solution should be integer. You will learn about integer programming in chapter 8.) ‘The net contribution of all clothing items equals $840,000 + $1,226,00 + $ 2,025,000 + $2,771,933.33 = $6,862,933.33. So far we have not considered the sunk costs for the three fashion shows and the designers which total $8,960,000. The total profit equals $6,862,935.33 - $8,960,000 = -§2,097,066.67. So, TrendLines actually loses almost 2.1 million 4-64 ©) If velvet cannot be sent back to the textile wholesaler, then the whole quantity will be considered as a sunk cost and therefore added to the fixed costs. The objective function coefficients of items using velvet will no longer include the material cost. The objective function coefficients of the velvet pants and shirts are now $175 and $40, respectively. ee r x aio" vale FT{['souten| 4200 ‘st77.777778. S000 sees.aeseey 6000 15762.06208 fa|-fatniman| 4300 geo | S000 c ° ra 13] {usaram| 000 —| 20000 | —so00 | — ssa | ane — | 9008 ‘The production plan changes considerably. TrendLines should produce 4200 wool slacks, 3177.77 skirts, $000 wool blazers, 3666.67 velvet pants, 6000 velvet shirts, and 15762.92 button-down blouses. The production decisions for all other items are unaffected by the change. The net contribution of all clothing items equals $840,000 + $1,226,00 + $ 2,025,000 + $2,983,822.22 = $7,074,822.22. The sunk costs now include the material cost for velvet and total $9,200,000. The loss equals $9,200,000 - $7,074,822.22 = $2,125,177.78. 4) When TrendLines cannot return the velvet to the wholesaler, the costs for velvet cannot be recovered. These cost are no longer variable cost but now are sunk cost. As a consequence the increased net contribution of the velvet items makes them more attractive to produce. This way the revenues from selling these items can contribute to the recovery of at least some of the fixed costs, Instead of zero TrendLines produces now 3666.67 velvet pants. These pants also require some acetate and thus their production affects the production plan for all other items. Since itis not optimal to make full use of the ordered \elvet in par (b)itcomes as no surprise hat the los npr (cs even bigger han i part ). ¢) The unit contribution of a wool blazer changes to $75.25. re el TrendLines should produce 4200 wool slacks, 10066.67 skirts, the minimum of 3000 wool blazers, no velvet pants, 6000 velvet shirts, and 6577.78 button-down blouses. ‘The production decisions for all other items are unaffected by the change. The net contribution of all clothing items equals $840,000 + $1,226,00 + $ 2,025,000 + $2,436,933.33 = $6,527,933.33. The loss equals $8,960,000 - $6,527,933.33= $2,432,066.67. 4-98 f) The right-hand-side of the acetate constraint changes. —— a 4200 _avaa333931___ 5000 ‘eo "geo | ato TrendLines should produce 4200 wool slacks, 14733.33 skirts, the minimum of 5000 wool blazers, no velvet pants, 6000 velvet shirts, and 355.55 button-down blouses. The production decisions for all other items are unaffected by the change. The net contribution of all clothing items equals $840,000 + $1,226,00 + $ 2,025,000 + $3,490,266.67 = $7,581,266.67. The loss equals $8,960,000 - $7,581,266.67 = $1,378,733.33. 4-71 8) The net contribution of one cashmere sweater sold in the November sale equals 0.6%$450 - $150 1.5*$60 = $30. After producing 4000 sweaters to be sold in September and October TrendL ines has 3000 yards of cashmere as raw material left. It is optimal to produce 2000 more sweaters using the remaining 3000 yards of cashmere yielding an additional contribution of 2000830 = $60,000. For the three linear programming problems determining the production plans for all other clothing items we need to include new decision variables representing the number of clothing items that are sold during the November sale. Clearly TrendLines does not want to produce items with a negative net contribution. Therefore, we need to consider only those clothing items that have a positive net contribution after taking the sales price into account. AT_= c 5 E Fé i 7 [z 1 3 ‘Activity a] — [Constraint [sik biouse [sik camisole Constant RS 5 st [7] [ant pa — S| | Solution | 7000 115000 ° [S| —~| Maximum} 12000 | 15000 | 36000 _ ~ Bf D E £ t aT n 1 2 ae re = 3 — 4 | | Gonataat Consiaint Ris] S| [woot 15 15 os | 05 30000 6) ip i i =i i 0 unit profit] 66.25 7425 T3376 | 878 |a02s000 Catt sauion | © ° ‘0000 8 | i It only pays to produce 2000 more Cashmere sweaters. The production plan for all other items is the same as in part (b). The sale of the Cashmere sweaters reduces the loss by $60,000 to $2,037,066.67. ofr 42 a) We define 12 decision variables, one for each age group surveyed in each region. Rob's restrictions are easily modeled a5 constraints. For example, his condition that at least 20 percent of the surveyed customers have to be from the first age group requires that the sum of the variables for the age group "18 to 25” across all three regions is at least 400. All his other requirements are modeled similarly. Finally, the sum of all variables has to equal 2000, because that is the number of customers Rob wants to have interviewed. 4-93 by ‘The cost of conducting the survey meeting all constraints imposed by AmeriBank incurs cost of $11,200. The mix of customers is displayed in the spreadsheet above. Sophisticated Surveys will submit a bid of 1.15*$11200 = $12,880. We need to include the new lower-bound constraint on all variables. fs0§14:$6 J$0$17:$6$17 >= $0$i9:5<419 sigi4:gig 16 >= $3414: $3816 is1§20 = $3820 ‘The new requirement increases the bid to $13,095.62. 44 4) We include upper bounds on the total number of people surveyed in Silicon Valley and from the age group of 18 to 25 year-olds. har eoge Saeed | = $0421 $17 >= $O$19:8G519 = $514 S16 >m $3$14:$3416 20 The new requirements increase the bid to $13,311.25. “9s ©) The three cost factors for the age group "18 to 25” are changed, aia | ets [ etme Tam ‘0 80800 400 600 200 80 | 950 72,038 pe [oe | oe fergaae7e 0 | 360 | ee ama nem smuawealei? ———/esiate ice) — With the new cost factors the bid increases to $13,828.75. 4-76 f) We eliminate all lower and upper bounds on the age groups and regions and replace them with Rob's strict requirements. These requirements also ensure that exactly 2000 people are surveyed so that we can drop that constraint too. oT Parana SUM er Rosin tctiy ~fortogc9 able preaete Rob's strict requirements increase the cost of the survey by $450. The new bid of, Sophisticated Surveys is $14,346.25. 499 ay a&b) Date Percantage Percentage Percentage Numer ot inet n 7th Bh —_—Bustng Cost (Studer) ween ___Studots___Grade__—Gease__—Grago_Sehoch 1 School 2_Sehool 3 2 00 7 9 = |B op 3 550 oa ose 08 | 600 300200 4 360 028 0 eae | 200 500 5 500 oa oe a? | 400 ‘ 450 ose o2e 0.38 | 500 _s00__o Capeciy. 90011001000 sotaton umber of Suse Asie Soot Seteol?__Scheold_— Toa! oa 7 7480 v a0 = 80 ‘rea 2 © 422,222002 177.777778| 600 00 ‘ead 0 227.7778 922.222222| 550 580 amas | 360 ° ° 360 350 ‘eas |ase.scess7 0 190,000909 500 500 ‘ama 6 [oas930099 0 _aee.seeee7| 450 = 480 Tout 600 71001000 capacty 900 1100 1000 Total Busing Corts $555,556.56 Consens: mn Graders [F69.930000 Bee 555556 G00. 11117 fn Graders [249.080067 969200939 360 20% of Tol 240 390 300 26% of Total 268 396 360 ©) The recommendation to the schoo! board is to assign students to schools as shown in the above solution section of the spreadsheet. Quantities that are not integers must be rounded since partial students cannot be sent. 478 4) The following solution decreases total bussing costs by over $135,000 but violates the grade constraints that were imposed. Solutions will vary and those than satisfy the grade Constraints will be likely to increase the total bussing costs. ate: Percentage Percentage Percentage Numero! inh n 7m Busting Cont (Student) ren ___Students__Gindo__Grage_—Grase_ School 1 Scheel? School 3 T @50 [0.52 0.8 03] 300 0 700 2 ooo | 097 ome 05 400500 3 550 03. ose 0.38 | 600 300200, 4 aso | 028 o oz | 200 500 5 soo | 039 © oa oar | 0 ol 6 aso_| 034 028 028 | soo __s00__o Tapaclyy 9001100 000 Sotuion Number of students Assones SscmooltSehool? __Senool3_—Toal en 7 750 . 450 450 ‘on 2 ° 600 ° 600 00 0a 3 ° ° 550 550 550 aan | 950 ° ° a0 = 350 fees | 500 ° ° soos 500 or ° ° 450 45000 F450 Toul 85010801000 caseey 9001001000 Total Buesing Cost = $ 420,000.00 rede Constr: School! __Seteol2__sehool 3 fan Graders [299 366 318 mm Gerdes | 310 338 302 fm Geagers |_2a7 365 300 20% of To) 265 315 300 20% of To 908 378 360 47 €) The number of students assigned from each area to each school changes to the solution shown below and the total bussing cost is reduced by almost $162,000. Date Percentage Percortago Pecerage Number of in @th in 7th nth —_—Buseng Cost (Student) Aroa___Students__Grade_—Grase_Graso_ School 1 Sehosk 2 School 3 i 450 a2 0.98 33] 900 0 700 2 600 a7 oats 400 500 3 550 03 sz 0.98 | 600 300 “ 350 028 vs oe | 0 00 5 500 eae oe = oar |i 0 400 6 450 De ere lene Capacity 9001100 T0905 Solaton: Number of Students Asie School Sohool 2 Scheold_—Total da 3 50 7 450 450 ‘wa 2 ° 00 ° 6008 = (600 nroa 3 ° ° 550 550 = 50 wena | 350 ° ° 380 380 woas [aresereie 0 ter.eserea] s00 = 500 raieiate2 so 26a.teiare] «50 450 Total 800 71001000 capacty 900 100 1000 Total Buesing Costs $393,636.36 rade ‘Consens: School! School? School 7m Graders |205,000009 263 312,900001 fm Graders |_240 366 360 0% of Taal 240 "a0 200 20% of Toa 268 396 360 f) The number of students assigned from each area to each school changes to the solution shown below and the total bussing cost is reduced by over $215,000. ate: Perceatage Percentage Percentage Number! inh nh Bh Busing Cost (Student) Suuderis Grae Gra Genoe_—Setool 1 Schoo? Senool3 rn 33] 0 0 700 soo | oar 02a = 0. | = «400500 550 03 os2 ose | oo 0 sso | 028 04 ose | 0 500 - soo | 039 ose o27_ | 0 400 aso_| 036 02a o8_| soo o_o Tapaciy, 9001100 F000 Number of stunts Assgnes SecrooltSehool 2 Seheold_—_Tota! a ToNeTTT ai 20a 0 450 450 © 296.859130 s62.<40861] 600 600 0 77.8609 472.0401] 650 = 580 360 ° ° sso = 380 jsasaae7t 0 ga.stesze0] 600 = 500 rs.s0a4si7 a74saasae 0 450 450 001400800 0014001000 Total Busting Costs $340,053.76 rade Constraints: Seroolt__Sehool 2__senooia fm Graders [— 208 Be. 780688 901.247919] Ti Graders| 324 952.247912 274752608 20% of Teal 270 330 270 # students walking more than 1.5 miles Cost # students walking to 1.5 miles current | $555,555.56 0 1 $393,636.36 0 2 $340,053.76 491 h) Answers will vary. 481

You might also like