precy MM ELT) d (ees Edexcel AS and A Level Modular Mathematics D) p A PEARSON COMPANY eesntvaneing lng changing bras py tol ital) Fh An sdant= ha [oc Edexcel AS and A-level Modular Mathematics Susie G JornesonContents About this book 1 -Bansportation problems 1.1 TRemninology used inidescsibing andl modelling the tansportation problem 12. Binding an initial sclution 40 the transportation problem: 13 adapting the algonthm 10 deal with unbalanced tmnsportation problems 14 Rnowwing what is meant by a deyenerate solution and how tomarage such solutions LS Birding shadow costs 16 Binding inprovernent inrtices and using these taind entering cells L7 Using the stepping stone method 10 obtain an improved solution 18 Formulating a transport problem asa linear programming problem, 2 Allocation @esignment) peoblems 2.1 Reducing cost mateices 2.2 Using the Hungarian algonthm todind a leas cost allocation, 2.3. Adapting the algorithm to use dummy loeation. 24 adapting the alponth to manage incomplete data 25 Modifying the algorithm to deal with a maximum profit allocation, 2.6 Formulating allocation peoblenes as linear pag ramening problems 3 The temelling silesman pcblem 2.1 Understanding the terminciogy used 32. Knowing the difference baween the classical ardl practical problems 2.3 Convertitny a netwerk into a complete netwcds of least distances 34 Usinga minisnum spanning tree method to find an epper bound 25 Using a minimum spanning tree method to find a lower bound 3.6 Using the nearest neighbour algorithm to find an upper boursl 40 Further linear programming 4.1 Formulating problems as linear progmamming pecblems 4.2. Formulating problems as linear programming problems, using slack variables 43 Understanding the simplex algorithm ta solve manimising linear prngramsning ‘pacblems 44 Solving manimising linear progsamming problems using simplex tableaw 45 Using the simpler tableau method tosolre maximising hnear peogramming problems seq tiring integer solutions Review Exemise 1 v oe 1s. vy 26 22 33 a4 43 44 48 $2 61 62 62 62 67 78 78 88 aa aL 4, wL ug BL5 Game they 5.1 Rnowing about two-person games and payoff matrices $2. Understanding what is meant by play safe strategies 5.3. Understanding mhat is meart by a zerosum. game $4 Determining the play sale stuategy for eath player SS Understanding what is meant by a stable solution Gadkdle point) 5.6 Beduring a pay-off matsix using dominanreagguments S.7 Determining the optimal mined strategy fora game with no stable soluuon $8 Determining the optimal mixed strategy forthe player with twochoices ina2x 30rd x 2game $9) Determining the optimal mixed strategy for the player with three choices ina? x 3or3 x 2 game $.10 Converting 2 x 2, 2% 2 and 2x 3 games into linear poogram ming pooblems Metwcrk flows 6.1 Knowing some of the terminology used in analysing lowe through nenocks 62. Understantling what is meant by a cet 63. idling an inital flow thicugh a capacitated dinerted setwonk 64 Using the labelling piorediure to firrl flow augmenting soutes tinesease ‘the flow through the nerwonk 65 Comfirming that a flow is maximal using the marimem fiow-minimum cut theorem, 6.6 Adapting the algorithm to deal with netsoceks with multiple somes and/or sinks Dynamic programming 7.1 Understanding the terminclogy and principlesc dynamic programming, including Bell man’s principle of optimality 72 Using dynamic programming solve maximum ard minimum problems, ‘presetriad in netmork form, 73 Using dynamic programming 40 solve minimax and maximin problems, presetriad in netmork foo, 74 Using dynamic programming to solve maximum, minimum, minimax or Tiustimin problems, presented in table soem, Review Exemise 2 Fxamination style paper Answers Index SEL Bs 6 Bz a7 Le 4 Ms 49 64 sz 16s. 166 wi we B2 3 we? nF me no 217 21 BER al mas.In this chapter you will th In industry, people are concerned with efficiency and cost-effectiveness at all stages. In this chapter you will look atthe costs dueto the transportation of goods - to factories and from factories to warehouses and customers. The problems considered are usually concemed with minimising distribution costs where eam the terminology used ta describe and model transportation problems leam about the 'northavest comer method’ leam about unbalanced transportation problems understand what is meant by a degenerate solution eam to find shadow costs find improvement indices use the stepping-stone method formulate a transportation problem as a linear programming problem Transportation pre oblems here are raultiple sources and multiple destinationsHAPTER 1 ‘V1 You should be familiar with the terminology used in describing and madelling the transportation problem. In ceder tosolve transportation problems you need to consider The capacity af each of the supply polnts (ar sources} - the quantity af goads that can be produced at each factory ar held at each warehouse, This i3 called the supply or stock, The amount required at each of the demand polnts— the quantity of goods that are needed af each shop or by each customer. This is called the demand (ardestInation) m Thauntt cost of transporting gaods fram tha supply points to tha demand points ae The unit cost & thecost of transporting ona item. fone item costs epounds 10 transport from Ato % then two tems will cost Ze pounds to transport along that muta, anda tems ne pounds Three suppliers , Band C, each paoduce oad grit which bas to be delivered to council depots W, X, Vand Z, The stock held at each supplier and the demand from each depot is known. The cost, in pounds, of tansporting one Ioery load cf gait from. cach supplier teach depot is also known. ‘This information is given in the table This bk & often referred to as the cost matri. Thic 6 the cost of ‘transporting one lon load from & 10 ¥ (ins) “ Notice that the total supply & equal to the total demand. If thir & not the care we simply introduce a dummy destination to absorb the. excess cupphy with transportation cost: all zero (see Section 1.3) Use theinioemation in the table toweite down a the numberctf lowy loads of grit that cach supplier can supply br the numberctf loury loads of grit required at each depct © the cost of transporting @ lorry load of grit from Ato WE 1 the cot of transponting 4 lotry load of gat fiom C 102, © Whithis the cheapest soute to use? £ Whichis the most expensive somte to use?‘Transportation prablems Suppliers A, B and ©.can provide 14, 18 and 20 lorry bads ms pectively Depots IN, XYand 2 requite 11,6, 44-and 42 brry bads res pectic ty The cost of trans porting ane lorry laad fram Ato Wis S180. The cost of trans porting ane lorry laad from © to Zs B20. The cheapest route B Ato x at £710 per bad. The most expensive route & A to Z at 2250 per lad |p tle lale Solving the transportation problem “The method is as follows 4. Fst find an initial solution that uses all the stock and meets all the demands 2, Calculate the total cost cf this solution and see if it can be reduced by transporting some geodsalony a route not currently in fhe solution, (Uf this is not possible then the solution is optimal.) 3 Lethe cost can be reduced by using a new soute, as many units a3 possible are allocated to thisnew route to create a new solution. 4 The new solution is checked in the sume vay as the initial solution to seeif at is optimal. Ténct, any new routes found are included. 5 When no fusther savings ae possible, an optimal solution has been fourd. 2. You can find an initial salutian to the transportation problem A method often called the ‘north-west corner method’ is u=d, 1 Create a table, with one rom for every source and cnecolumn for every destination, Each call represents a route from a source toa destination. Each destination’s demand is given at the font cf each eclumn and each source's stock is givenat the endl of each sow, Enter numbers int each cell to show how many units are to be sent along that routs, 2. Begin with the top left-hand comer (ihe necth-west comes), illocate the maximum available quantity to meet thedemardl at thisdestination (whilst not exceeding the stock at this source!) 3 Ascath stock is emptied, move cre: square down and allocate as many unitsaspossible from thenext source until the demand of the destination is met. seach demand is met, move one square to the right and again allocate as many units as possible. 4 When all the stock is assigned, ard all the demands met, stop Inondar to avoid degenerate Fora problem involving Inthe exanaination, solutions, movements ame mice msoume vertices andre pooblens will be between squares either vertically destination vertbes yom rust restricted a orhorkontally but never enterm tra — Ltransportation maimumied 4supply diagonally See example 4 quantities & (0, This will points and 4 demand. not reduce thmughout the pains, pooblern.HAPTER 1 Depa W | Depo | Depary | Depoz | Sock Supplier A 180 110 130 280 4 Supplier B 120 250 180 280 16 Supplier ¢ [240 270 180 120 wD Demand i 1s ie 0 80 2 Beginning vith the north-west comer, start to fill in the number of units you wish to send alang each route ‘Use the north-west comer method tofind an initial solution to the problem described in: evample ] and shown in the table. 1 Set upthe tabke 2 Start to fill in the number of units you with to send along each route. Depot W requires 11 lorry oad, This does not echaust the stock of sup plier A 3 The demand! at WV has been met 20 moveore squnr to the right and allocate 14 - 11 = 3 unte The stock at a new ethawted. Thedermand at X has not bean met As the stock at Aas been exhausted, move one square clown and allocate the maximum possible number of units tron sup plier 20 chepot K. In this cace, 15-3512.‘Transportation prablems Now that the demand at X har ako been met, move one square to theright and use the rerraining stock at & to stort to met the demand at, The stock at B & now exhawted (12 + 4= 16) and so move onesquare down and use thestockat to tuff the remaining errand at ¥, Finally, move one square to the right and use the remaining stockat to mest the demand at Z This is the final table. Allof the stock has been used and all of the demands met. The number of occupied cells (mutes wed) in the table = numnber of supply paints + numberof cemand poin’s ~ 1 In this ease the rmumiber of occupied cells (routes wed) = 6, mumnber of supply points = number of demand points = 4 amlé=8+4-1 Use this table, together vith the table showing costs, to work out the total cost of the solution The cel shaded indicate the costs of the runes curently being used. tie thers, together with the number of units being transported along 26th mute, that give ‘the-cest of the solution, The total cost of this solution (NX IED) + BX NOV + (2 x BOO) + (FX 10) + (10% WO) + 10 x 129) = E9000,HAPTER 1 1.3 You can adapt the algorithm to deal with unbalanced transportation problems. When the total supply > total demand, we say the prablem is unbalanced. if the problem is unbalanced we sim ply add a dummy demand point with ademand chosen sa that total supply = total damand, with transportation costs of zaro a] 2 [© | supply x 8 ll [io 40) ¥ w [8 12 60 iz 2 fF 8 s0) Demand [ so | 40 [ 30 Three outlets a, Band C ate supplied ty three suppliers if, ¥ and Z. The table shows the cost, in pounds, of trans pceting each unit, thenumber of units requiredat each outlet ard the number cf units available at each supplier. a Explain why itisneressiry to add a dummy demarrl point in order to salve this problem. b Add a dummy demand point and appinpriate ccsts to the table © Ube the northwest comer method to obtain an inital soluucn, | a The total supply & BO, but the total demand is 20. A dummy needed to absorb this excess, s0 that total supply squale total demand. Weadde dummy column, D, whem thademard & 20 (the amnount by which the supp by eicaers the cerrand), and ‘thetrane portation costs are gem (since there & na acta ‘transporting done}! The problem is now balanced, the total supply = the total errand‘Transportation prablems manage such solutions 1A You understand what is meant by a degenerate solution and know how to @ In afeasible solution to atransportation problem with mrowsand n columns, if the number of calls used is less than a+ am — 1, then the solution i; degenerate. This veil happen when an entry, other than the last, is made that s esthe supply fora given rove, and at the sametine, satisfies the demand for a given colurnn, Thealgorithm requiresthat n+ m—1 cells are used in every solution, so a zero needs to be placed in a currently unused call a Demonstrate that thenocti-west comer method givera degenerate solution and explain why itis degenerate. Bb Adapt yoursclution to give a non-degenerate initial solution and state its cost a A Supply We 50. 50. x 20 20 i zo [1 | 26 z Eyes Cemand | 20 | 40 | 50 | 120 This solution is degenerate since it- fulfils all the supply and demand needs bur only uses 5 cells WA, XB, YB, YG and 20. There are 4 rays and 3 columns $0 a non degenerate solution will wse¢ + 3 Scale. ke Start by placing the largest possible number in the northwest corner & Supply 20 We x z Gemand | 20 | 40 | so Notice that ther: has been adliagonal ‘reo ve! frome cell Watto call XB. Degenerate solutions can be avoided by not allowing diagonal moves Having placed the 30) in the NW comer, you now need to place the next number in the square to its right or in the square undementh, Since both the supply and the derrand are satified, thir Taare that you will have to phoea zen in eitherthe call ie right or the cell urciemasth,HAPTER 1 There are bw possible intial sombions, depending on where you chose to place the zero Either or In fart the ze can be phoed, anywhere in the tab, but it is convenient to ‘stick to. the rule! shout restricting the movement to one square down or one square right Both hae a cost of (20 X 12) ++ O+ OX 5) + (20% 8) + (19% 7) + BX B= E880 Photocopy masters are available for the questions in this exercise. In Questions 1 to 4, the tables show the unit costs of transporting goods fom supply points to demand points. In each ease: a use the north-west comer method to find the initial solution, bi vendiy that, for each solution, the number of occupied cells = mumber of supply paris + numberof demand points - 1 © determine the cost of each initial salution, 1 PB Q R Supply A iso [213 [ 220 32 EB 17s | p04 [ 218 44 c ise | se [246 34 Demand | 28 | 4s | a7 110‘Transportation prablems Q Supply A a7 [24 Te 123 B is [21 [2s M43 c i [22 | ie 84 D ao | a7 | 16 150 Demand 200 | 100 | 200 S00 A sé [a6 [30 [a1 134 BE sa | 7a | ve | ts 203 c oz [mm | s7 [a 176 D so [aa [vs [71 187 Demand | 178 | 175 | 175 [ 175 700 a [2 [ c¢ [> | soppy x a7 [33 [ 34 [41 60 ¥ 31 | 29 | 37 | 20 60 z 4o_| 22 | 28 | as 80 Demand [ 40 | m [ so [ 20 Four sunciwich shope 4, B, Cand D can be suppliel with bread fim three bakeries, X, ¥ ‘and Z, The table shows the cost, inpence, of transpoming one tray of bread fom each, supplier to each shop, the rusher of trays of bread seq wired by each shop and the number ‘of tmays Of Dread that can be supplied by each bakery. a Explain why itisneressury to add a dummy demarsl point in order to solve this problem, and what this dummy point means in peactical terms. 1 Ube thenonth.wwest comer method ta determine an initial solution to this problem and the cost of this solution, Acompany needs to-supply seady snined concrete ftom fourdepots a, B, Cand D to four ‘work sites R, L, 2M and N. The number of loads that can be supplied from each depot andl the num ber of loads sequised at each site are shown in the table above, as well as the transportation cost per load from each depot teach work site. a Explain what is meant by a degenetate solution, b Demonstrate that the north-west comer method givesa degenerate soltstion. © ddapt your sdlution to give a non degenerate intial solution.HAPTER 1 7 L[™ [om [ supply P zfs? 22 2 a[3[7 a R é [4 [é@ 1 3 af2ts 3 Denand | 1s [ 17 [ 20 ‘Ter table shows balanced transpovtation problem, The initial solution, given by the noth. ‘west ecener method, is degererate @ Lee this infomation to determine the values of @ and b. 1b Hence write down the initial, degenerate solution given by the north west comermethod, Finding an improved solution To find an improved solution, you need to 1 use the non-empty calls to find the shadow costs (382 Section 1.5) 2 use the shadow costs and the empty cells to find Improvement Indices (see Section 18) 3 use the improvement indices and the stepping stone algorithm to find an improved solution (522 Section 1.21 ‘1.5. You can find shadow costs. Hm Transportation costs are made up of two components, one associated with the source and one with the destination. These costs of using that route, are called shadow costs. Tn scample 2, the cost of 2950 an transporting one unit fron supphier Btodepor X must be Geperrlent on the features — location, toll costs ete, of bot B and Tsing the rowtes curently in awe you can build up equations, showing the cost of trans porting ore ‘nit, suchas 30) + DG) = 110 and SC) +DEZ)=120 ete where Si), D(X) are the costs dueto supply point Aand demand point % and soon, respectively. Yu neat a value for eachof the soume compcments arr] each, ; Of the destination components, You do not have sufficient Tou sas only eo king 2 he equations for a sclution (fiveequations and six unlncwrs) bat your cument solution, selative costs will do.‘Transportation prablems ‘Tofitd the shadowecsts, follow these steps, 1 Start with the north-west commer, set the cost Put the first source component linked withits cure to ze ‘0 26m, and set the destination component to campy the cost of the 2 Move along the row to any other nonempty squares transportation in roe 7 and find any other destination ecets in the same way 2 When all possible destination costs for that row have beenestablished, go to the start of the next row. 4 owe along this row fo any nioreempty squares and use the destination costs found earlier, to esablish the sourcecost for the row, Onre that has been done, find any further waknown destination costs, & Repeat steps 3 and ¢ until all soureand destination costs have been found Calculate the shadow costs given by theintial solution of the problem. given in eample Zand shown in the table. Initial solution (see page S) was bat x ¥ z Shogk A ia 4 & wia 16 Gi ol] ol 2 Demand] i [ 6 [ #4 [ o [ so Focus on the cogk of Che toutes being wed — the non ‘Remeraber to use the cost ee uate values not the number ef iter curmntly being Depot | Depot | Depot | Capes ‘trarepottedalong that route, took w fox ae Suppler A[ BO | 110 4 Supplier B 20 | o 6 Suppler © wo | eo | 20 Demand “ | 6 | # | | 50Guerin Putting @(A) to zero, from row Te get DON) = 180 “put SCA) = arbitrary and and DX) = 42 then solvethe equations SC) + Dey = 160and Find the mercaining shadoweos's by ‘walking rund! the curent 0 solution, noting the shadow costs you find mound the edge cof the table, and wing shadows costs found extier to find the memuining ones. Shadow costs 180 110 Depot Wi] Depot x | CeporY | Ceporz | Stock a Suppler A | BO tO id Supplier & 20 60 6 Supplier C 190 ao | 20 Demand tl 16 os 10 50 Now mous to Row 2. You now that Dg) = 10, soyoufind S(B) = HO and ure this together hence DEQ = (B+ OCF iste aet C0) Shadow costs 180 110 1a Cepor | Depokx | Depot | Depo Z | Shook a Supplier & | 180 110 14 Supplier & 20 EO 16 Suppler C 480 qo | 20 Dentand tl 8 a 10 50 Mowe to Row. eres ‘ou know that; BOY) = 10, sawe find SUC) = 180, and S(G) = 180 and hence that DZ) = 120 to fine O62). Shadow costs 180 110 1a =60 Depot | Deporx | Depot y | Depot z | Stock a Suppler & | 180 Wo 14 wo Supplier & 250 60 6 180 Supplier 6 180 zo | 20 Demand tl 5 + 10 50 YOU have now found all soure and all destination s haslow costs. SA) =O SB)=4O sc) = Be 5: npc asians Do not bealarmed at the negative shadow DW) = BO DR) =O Pt) = cost found tor XZ). erent DZ = -é0 arbitrary putting 5(41 = 0. You are simply fincling costs relative to the cost SC.‘Transportation prablems Calculate the shadowccets given by theinitial solution of the preblem given in ecample 3 and shown in the table, The northavest corer method gave the following intial golurion (Gee page S) A [8 [¢ [2 | Suppy x 40 40 ¥ % | 40 [6 60. i 20 [zo | x Demana [50 | 40 [20 [50 We need to use the costs rather than the number of items being thang ported. So we use the following mumbers. Shadow ooste ate [ic [oo | supp i 2 aa ie ole le 60 Zz @ | o | Demand | so | 40 | 30 | ee [ Bo Arbitrarily assign 20) = 0. Shadoweosee a |e [ec [2 | supp ° x 3 40 i © [a [2 $0) z a [ol % Demand | 50 | 40 | 30 | eo [ Bo Use this to work out the shadow cost for (AY Shadowaosrs 3 A Supply x 2 4. ¥ |e | 2 a] Zz @ | o | 50HAPTER 1 Use this to work out the shadow cost for S(V). Shadowaosrs 3 ate [ec |b | supp o x 2 40 1 i ola |e 60 Zz @|o|[ = Demand | 50 | 40 | 30 | ae [ Bo We use this to work out the shadow costs for D(B) and D(C) Shadowaosts aft? [ Ae Te Te | suppiy x 2 40. 1 Y 10 12 a] z e |o | 0 Demand | 50 | 40 [30 [ea [ Bo Use these to work out the shadow oost for (2) Shadowsosts as fri A B c D_ | Supply a x 2 Ea 1 i ol] a | 2 60 = z a flo Ea temana] 30 | 40 | 20 [oT 0 Use this te work out the shadow cost for D(D Shadowaoses 2 [7 [" 3 a fe Te [2 | supp o x 2 40 1 ¥ ole |e 60 =3 @ a[ol Demand | so | 40 | 20 | 20 | 50 You do not have to show each stage of the table in the examination, Just this fal list of shades costs i sutfcient,‘Transportation prablems 1,6 You can find improvement indices and use these to find entering cells Tt may be possible toreducethe cog of the initial solution by intioducing a romte that is not currently in use, bu consider earh unesed soutein tum ard calculate thereduction in cost which ‘would be made by sereling one unit along that route, This is called the improvement index. Wm Theimprovamant index in sending a unit from a source Pte 4 demand point Qis found by subtracting the source cast §(P) and destination cost (0) fram the stated cost of transporting ane unit along that route C(PQ).i2 Improvement index far PO = lkq= C(PQ) - 8(P} - D(a) @ Theroute with the most negative im provement index will be introduced into the solution The call corresponding tothe value with the most negative improvement index bacamas the entaring call (or antaring squara orantoring routa} and the route it replacas is raferrad toastheextting call (oraxtting square or axtting route). Wm Iithere are two equal potential entering cellsyou may choose either. Similarly, if there are turo equal exiting cells, you may select either It there are no negative imprevement indices the solution is optimal. a ee | [rena | Sep [Bop [Spot | so | memati fm fe Ps [a | Peeps [pe [ae [ise oe 180 Ue theshadow costs found in example §, and shown in the table above, tacalculate improvement indices, and use these tosdentify the entering cell Foous on the routes not currently being used, BN, ON, CX, AY, AZ and BZ You already know that BA) =O SB)=140 SC) = BO DW)= BO DR=to PM=27 P= -60 Improvement index.for BW = |, = (BN) - S(B) - IW) = 190 - Ha - 180 = —180 Improvement index for OA = by = 240 - 160 - Bo = —tao Improvement index.for OX = be = ZAO — 180 20 Improvement indexfor AY = ly = 180-2 —10= 20 Improvement index for AZ = 2 = 220 - 0 - (-80) = 260 Improvement index for BZ = Ibe = 289 — 149 — (80) = 200 The entering cell « therefore BM, since this © the most negativeHAPTER 1 a Lie thenorth.mest eceer method to find an initial sclution to the transportation problem shown in the table b Find the shadow costs and improwement indices © Hence determine if the solution is optimal 6 There are no neqatice improvement indices, so the solution is optimal, Photocopy masters are available fox the questions in this exercise. Questions 1 t04 Start with the initial, north-west comer, sclutions found in questions 1 to 4 of exesrise 1A, Tn qach case nse the initial solution, and the original cost mate, shown below, to fd a the shadow costa, Br the inprovernent indices © the entering cell, f appropriate,‘Transportation prablems 1 Pye |B | supply A 150 | 213 [ 222 [32 B 17s | 204 [21a [44 c ise | 198 [ 246 [34 Demand | 28 [ 45 | 37 2 B Q | R | Supply A i [ 24 [og 122 E is [21 Mz c ia [22 [is 84 D ao [a7 | 16 150 Demand | 200 | 100 | 200 ‘7 You can use the stepping-stone method to obtain an improved solution. Inecample 7 you distomered that the most negative improvement index was BW with a value cE 130 Gee page 15), This means that every time you senda unit along BW you save a cost of 130, Therefore pou want +0 sendasmany «nits as possible along this new route, Sou have to be careful, however, not to exceed the stock or the demand, To enewe this, you go thoough a saquerre of adjustments, called the stepping-stone method. im are looking therefore fora eyele of adjustments, where you increase the valued ore cell and thendecrease the valucin the nent cell, then inceasethevalue in the next, and soon, A popular mind picture is that yousme using the cel as Stepping-stones', placing one foot on each, and alternately putting down your left foot (increasing) then right foot Gecressing) a you joumey amund the table —hence the methad’s nickname,HAPTER 1 The stepping-stone method 1 Cratethecyele cf adjustments, The tmu basic rules are: a Within any iow and any column therecan only beone increasing cell ard one decreasiny cell I apart from the entening cal, adjustments are only made to non-empty cells, 2 Once the cytle ofadjustments has been found you transfer the maximum tumberof units through this cycle, This will be equal to the smallest number in the decreasing cells (since YoU may nck have negative units being trarsportesd) Xiu then adjust the solution toineorporate this im provement In sample 2, the table of costs was ata cost of £3010 (ee page 5) ‘Cbiain an improved solticn and find the improved ecst Use BW as the entering cell, since thie gave the most negatwe improvement index, 130 (see Example 7, page B) So EW will be an increasing cell Wie enter a value of into this cell w [ox | y | Zz | Seok A fies 6 @ | 12 4 16 20. 7 a Ria] a3 Pemand] 1 | &‘Transportation prablems In order to keep the demand at Woorrect, you must therefore decresse the entry ab All, so ANS will he a decreas ng cell wofx Tye | stock A ia 3 a 6 e[el4 16 a ool 2 Cemand| a [oe | 4 [ol so Inorder to keep the stock ab A correct, you must therefore iltcheage the entry at AX, 2o AX will be am mtereas ing cel Ww. x vy | z | Steck A i-@[3+¢@ 4 6 @ @ 4 16 a ole] 2 Pemand | 6 |wi[wo] Inorder to keep the demand at X coract, you must therefore decrease the entry at BX, 2o BX will ke a decreas ing cell ut ae ae A i= @[3+e + & @ [eel 4 1 iG ol o | 2 Demand ie ara | os | Nowechoose # value for @ the greatest value you oan, without in-roducing negative entriss inte the tabls. Look at the decreasing cele and see that the greatest valur of 8 is H (since 1 — 11 =O} Replace d by tin the table: Ww x vy [2 | Stock & wan sen 4 B Ce 16 c wo | 20 Bemand | tt s [ui[wo] soHAPTER 1 proved solution: a ae a A i 4 6 1 4 16 a 0 | 0 | 20 Bemand [tl 1 [# [oe] 2 Locking at the table of transportation costs, this solution Ue lereae aAa 12) Asa double check, ite alvmys true that New cost = cost of formersolution + impmvement index % In thi case 7580 = 9010 +(-120)x 11 T ‘bu will notice that JAM has become empty. AW is thereines the exiting cell, Remember that the number of cells tised ina feasible solution must equal thenvumber of tows lus the numberof echumns mitras 1. Saif you pat a number into an entering cell, dt must be balanced by a number being removed fom an exiting cell, At cach iteration we create one entering cell and one exiting cell H Tofind an optimal so lution, continue to calculate new shad aw costs and improvement indices and than apply tha stapping.stone mathad, Repaat this tration until all the improvement indices are non-negative. Find an optimal solution forexam ple 9. ‘This second iteration indivate: the amount of working pouneedto show in the exarnination, Second iteration Find the neve shadow costs: Shadowsosts: 60 | te | © |-ao w fx | y | z [ stock o 4 wo] to [io [zo] 4 HO 8 190 | 260 | 150 | 220] 16 160 fe 240 [270 | 190 | eo | oo Demand] 1 | 6 | # | 0 50.‘Transportation prablems Finding the new improvement indices for the nen-used cells: Ali = 180-0 -50= 80 OW = 240 — BO - 50 =10 % = 270-120 - 119 = -20 WO-0- = 220 -0 +0 = 250 220 — 40 + 60 = 200 " AZ B So che new encering cell BOX since thes has the most negarae improvement: index. Applying the stepping-stone method gies te x ¥ z_| Stock A 44 haa & n_|i-e@|4+e 46 c 9 |w-a| o| 2 Pemand | 11 6 4 | ol so Looking at celle BX and Cv we see that the greatest value for ¢ 61 The new exiting cell will be BX @ = 1 and we get w | x | vy | z | Steck A 4 + & tt 5 16 c i1[e [eo] 2 Penand] 1 [oe | 4] eo [ so The new costs 27560 ‘Cheeking, 7580 + (-20) x 1 = 7560 Third iteration New shadow costs: Shadowsoses 72 | no | 20 | -40 wef |v [|Z | Stock a 4 wo | no | 1 [seo] 4 yen B 190 | 0 | 150 | 220] 16 180 c 240 | 270 | 190 | 120 | 20 Demand] a | 6 | # | 2 EyHAPTER 1 New improvement indices for the norrused cells: Alli = BO — 0-70 = 110 CW = 240 — 160 - 70 = 10 80 — 120 — 10 = 20 0-0 - 30 = 100 290-0440 = 200 220-20 + 40 = 200 There ate no neqathe improvement indices #0 this solution i optimal. The solution & 110 unite A te x 190 units B tow ‘At this point, if ther isan improvement index ot 180 units BtoY 0, this Wowk incitare that there is an aemative ‘optimal solution, To find, simply we the cell 27D unite © tox swith the 2am impmmvernent index as the entering 190 units ¢ toY all, (ee question 2 in hibeed Buemise 1B) 120 units Cto Z Some stepping stone noutes are not rectangles and some 4 values are not immediately apparent Supermarket % | Supermarket ¥ | SupermarketZ [Stock a 22 28 13 26 26 in iL ‘Tee table stows the unit cost, in pounds, of transporting goods from each of theee warehouses, A Band C toeach of these supermarkets X, TandZ, Italo showsthe stock at each warehouse and the demand at each supermarket Solve the transportation problem shown inthetable, Use the northeest comer method to cittain an initial sclution You must state your shadow ccets, improvement indices, steppiny AONE Moss, @ values, entering calls and eating cells, You must state the initial cost and the improved cost after each iteration. Check the problem is balanced. Supply = Demand = 38, sowe de nt nesd to add a durin. The northwest corner method gives the sollowing initial solution x [ov | z | Steck A 0 3 15 6 im ie tt c ta ®‘Transportation prablems The cost of the initial solution it £220 Calculate shadow costs Shadowsosts eA | 2 | £ x | vy |Z | Seok o 4 e4 [ee | ce | 13 + B 26 | 2 | 14 1 10 a 20 | 22 | a 2 Demand] | 6 | 8 Caloulating the improvement: indises for the empty cells: BX = 26 - 4-24 = -2 OX = 20-0 24 = 14 cY=22- 0-22 = -10 AZ= 2B -0- = 18 Use OX.as the entering cell, since it has the most negatke improvement index. ‘This stepping stone zoute is quite complicated Take aminute or wo tocheck how it har been erated Start by putting #in CX, then to comect the dermand of X, subtract (from cell AK, then 10 comect the supply in #, add to call AY and so on, finishing at cell C2, The maximum value of # i 42 Either cell AX or cell BY can be the exiting cell, since both of these will go to zero, We simply cheose one of them, AX to be the exiting cell, the other vill have: a numeral value of zero ‘Many candidates fail to understand the ditference between an empty cell ancLone with a zero entry. 2am is a wumber, just like 4 and 'counts' towards our m +2 — 1 entuies in the table, An empty cell has no number in it, Improved golubion & x [vy [Zz | stock 4 6 1 6 o fl ti if 0 iz 2 Bamana | 12. oeHAPTER 1 The cost i now 2680 We need to check for optimality, ky caloulabing improvement indices The second set of shadow costs are: Shadowsoses xo | ze x | vy |Z | Stock o 4 24 | ez | a | 18 4 6 26 | ee [4 1 10 fe 2o | 22 | a 2 Demand | © | 8 | 8 Improvement indices ax [ee [cy | az 4 12 So the solution is not- yet optimal and the nest entering cell is CY. BY nd Shark A 6 Ge 6 o-¢@| +e au C wD a aw 2 Demand | 10 6 B We can gee from cell BY that the maximum value of #6. The entering cell is CY and will have an entry of 0 The ceiting cell » BY and a vill now be empty. The improved solution x v Shor k A 6 1B c iol om Ie: 8 pemand [| 42 | 8 | The cost i unchanged at 2680 We again need to check for optimality, by calculating improvement indices, Shadow oosts: Shadowaosts ao | zz | a a | vad zal |Rsrock: o 4 24 | 22 | | 15 el B 26 | 2 | 0 a fe ao | 22 | 20 # Demand] wo | 6 | 8‘Transportation prablems Improvement: indices are ax_| ex | sy | az 4 wz | 8 Allthe improvement: indices are non-nagathe and so the solution is optimal x Ly LZ | sek 8 8 1 6 n tl Gi olol 2 ® Bemand | © | 1 | Go the optimal solution is to send 1% units from Ato ¥ Hunits from Bto Z 10 units from ¢ to X Zunite from € to Z ab a cost of £680 Questions 1 102 ‘Complete your solutions 10 the transportation problems fiom questions 1, 2 and 4 in exercises Li and 1B. You should demonstrate that your solution is optimal