DECISION SCIENCES Dr. RANJEET H. CHITALE e PRAKASHAN k eo Scanned with CamScanner4 Book OF DECISION SCIENCES For M.B.A. Semester - II & Other Management Courses As Per Revised Syllabus MSK Choice Based Credit System and Grading System Dr. RANJEET H. CHITALE B.E. (E &TC), M.B.A. (Materials, Finance, Marketing) Ph.D. (Operations Management) Associate Professor Department of Management Sciences (PUMBA) S.P. Pune University (Formerly University of Pune) Price % 400.00 NIRALI PRAKASHAN EON aos 2979) Scanned with CamScannerSyllabus... Ll 11 12 13 2. in 2.2 3.1 3.2 3. w 4. m 4. uv = 5.2 Number of Sessions Introduction: (942) Importance of Decision Sciences and Role of Quantitative Techniques in Decision- Making. Assignment Models: Concept, Flood's Technique/Hungarian Method, Applications including Restricted and Multiple Assignments. Transportation Models: Concept, Formulation, Problem types: Balanced, Unbalanced, Minimization, Maximization Basic initial Solution using North West Corner, Least Cost and VAM, Optimal Solution using MODI. Linear Programming: Concept, Formulation and Graphical Solution. Markov Chains and Simulation Techniques: Markov Chains: Applications related to Management Functional Areas, Implications of Steady State Probabilities, Decision-Making based on the Inferences Monte Carlo Simulation, Scope and Limitations. Decision Theory: Concept, Decision under Risk (EMV) and Uncertainty. Game Theory: Concept 2 by 2 zero sum game with Dominance, Pure and Mixed Strategy. Queuing Theory: Concept, Single Server (M/M/I, Infinite, FIFO) and Multi Server (M/M/C, Infinite, FIFO) CPM and PERT (6 + 2) Network Calculations: Calculating EST, LST, EFT, LFT, Slack and Probability of Project Completion. Sequencing Problems: Introduction, Problems involving in Jobs-2 Machines, n Jobs-3 Machines and n Jobs-m Machines; Comparison of Priority Sequencing Rules. Probability: (6 +2) Concept, Addition, Conditional Probability Theorem based Decision-making; (Numerical based on Functional Areas of Business expected) Probability Distributions: Normal, Binomial, Interval Estimation Standard, Errors of Estimation. (8 +2) (6+2) Scanned with CamScannerIntroduction to Decision Sciences (unit - 1) 4. Linear Programming (Unit - 2) 3, Assignment Models 4, Markov Chains and Simulation (Unit - 2) 4a- 3) 5.4- 5. Decision Theory (Uni 6, Game Theory (Unit - 3) 7. Queuing Theory (Unit - 3) TA 8, CPM, PERT and Network Calculations (Unit-4) —” 8.4 -| 9. Sequencing Problems (Unit - 4) 94 - 410. Probability (Unit - 5) 10.4 — 41, Probability Distributions (Unit - 5) 14.4 Multiple Choice Questions Mca.1 - ‘Appendix A ‘Appendix 8 ‘ Appendix ¢ LINEAR PROGRAMMING PART I~ PROBLEM FORMULATION 1.1 Introduction 1.2 Linear Programming Problem (LPP) 1.3 Stages of LPP 1.4 Problem Formulation of LPP Solved Problems Exercise PART II - GRPAHICAL METHOD OF SOLUTION, 1.5. Introduction 1.6 Grophical Method of Solution 17 Note Solved Problems Exercise - I MTS 20:0 Ole es PART | - PROBLEM FORMULATIO! Linear Programming is a deterministic model in decision sciences. Ths isan optimisation techrique whereby we consider 2 given business problem and identify the variables conceming the same. These variables are then used to write down the objective function and) the constrains forthe given problem station. Under LPP both these are expressed using the Linear Relationships only. After the formulation ofthe given problem we techniques Tike Graphical method and Simplex method to solve it further, Problem formulation is the eg fram eee ost critical step under LPP model. nae : a Scanned with CamScannerLUNEAR PROGRAMMING is proportional ie. the power of the terms raised to 1, eg. P = ax + by + Cz, where a, b, les v rnonly faced by management is to decide the manner in rt se used to achieve the desired objectives like (6) Otjecive Function: This represents the mathematical equation of the major objective of the system in terms of the decision variables” oll (©) Linear Relationship: Linear Programming deals with the problems in which the each variable appears 3s have two properties: ea Function) (Quadratic Function) P = ax + by +c (Cubic Function) o an Condition: i.e. the decision variables must be either 0 or +ve (@) Feasible Solution: A set of values ofthe decision variables which satisfies all the Constraints and the non- ition is have may fstle scones nation a feasbe solution. A problem can () OptimayOptinum Solution: Th P : feasible solution which bectvefuncton. Normally an optimal sluton isunique, ns” *Ptmises_ te t/ctlem derttication trough colection po formulation ie ane to te relies formulate the mathematical problem from the given data. oe cae DEQSION sciences LUNAR pRocRANeING (Objective Function) Constraints (non-negatvity condition) contribution to the are the structural = 1 to mj= 1 ton) ‘and Fi. 1 unit of food F; contains F, contains 6 units of vitamin A B, Lunit of F, and F ind £5 respectively, The minimum daily requirement (fora person) of vitamin A and B is 80 and 100 units respectively. Assuming that anything in excess of the dally minimum requirement of vitamins A and B isnot harmful, find ‘out the optimum mirture of foods F; and fF; at the minimum cost which meets the daily requirement of vitafnins A and B. Formulate this as a LPP. 1 Scanned with CamScannerLINEAR PROGRAMING DEcION sc the form of a table as: SON Sconces LUNean pROGRAMNANG . The given data can be expressed in pT Solution: The ot f units per unit of food) 1 ‘company makes three products X, ¥, Z which go through three Vitamins (No.0 : (per unit) ssemibly. The hours of department time required by each of Food A a each of the departments and the profit contribution of 4 3 Th : ‘ 5 Products Time Required per Unit Profit contribution F a (per unit) Tadao ] a Drill Lathe [Assembly Requirement 80 | x 3 3 8 9 TL The problem sto rd the quatties of foods F, and Fz 50 ast meet the daily minimum ¥ 6 5 10 rf roblem i to 1 4 2 20 ni 3 requirement of vitamins A and imum His. Available | 210 240) 260 the unknowns to be found out. Hence, Decision Variables: These Let x= the quantity (number of units) of F: required 14 the quantity (number of units) of Fz required. 4. Objective Function: The objective is to minimise the cost of food. Now, the cost of x per unit cost = 4x). Similarly, cost of xe units of food Fe is (5x). + 5x = Clsay)). Hence, the objective function is to {1s consti The minimum requirement of vitamin Ais 80 unis. The number of units of vitamin A contained in x; us food F; ar number of units of A contained in 1 unit of F,) i.e. (3x). Similarly, number its of A contained in x: ‘units of F; will be (6x:). Thus, total intake of vitamin A will be (3x; + 6x2) which must. beatles 80 units "3K, + 61280 v= (1) (Vitamin A constraint) ity, in requemen of tain Bis 100 units ; M+ 3x2 100 ~~» (2) (vitamin B constraint) (3) ce the rumber of unis x, and " a %2 Gan not be negative, he 20 (Non-ngatity conten) = eaiwetae ‘Thus, the LPP format will be, Minis ¢= Subjectto 34.7 612 280. (amin A constant) 4114 342 2100 (vtamin 8 consint) ae (Non-negativty condition) 1a 41 + 53a (Obj function - Cost Minimisation) ‘The marketing department of the company indicates thatthe sales potential for products X and Y is unlimited, but for Z it is not more than 30 units. Determine optimum production schedule. Solution: 1L._The objective is to maximise the total profit by selling units of X.Y and Z which are [produced under the given production time constraints and sales constraint. 2 raduct X produced roduct ¥ produced y~ Number of units of - Number of units of product Z produced 3. Objective Function: Total Profit P = 9x + 15y + 202) must be maximised. Thus, Maximise P = Sx + 15y + 202 4. Constraints: Dill Department Constraint: Total time required for driling activity ie (3x + 6y + 72) ‘must be less than or equal tothe time availabe ie. 210 Hrs. Thus, 3x + 6y + 72210 Lathe Department constraint: In similar manner, 3x + Sy + 42240 ‘Assembly Department constraint Similarly 8 + 10y + 122<260 OR dividing ‘throughout by 2 we get 4x + Sy + 625 130 (iv) Sales constraint: For product 2 the sale isnot likely to exceed 30 units. 2 2830 «. The LPP format is, Maximise P= 9x+15y +202 (Profit maximisation) Subjectto 3x + 6y +72 $ 210. (Drill constai 3x + Sy + 42 < 240 (Lathe constraint) ‘4x + 5y + 62 $130. (Assembly constraint) 2530 (ales constraint) xy.z20 (Non-negativity condition) Fr Scanned with CamScannerLINEAR PROGRAYM ang ymade garments shop wishes tO publgy * marathi and one English. The expectey ‘and 1500 people per advertisement 2 Marathi newspaper costs € 3000 and for an English ded not to place more than 20 advertisements in the can ebeepapr and wat to pace at est © averse," (8 English daily. The total advertisement budget is © 50,000/- Formulate the problem as a LP model. Solution: Let x, ~ Number of advertisements tobe placed in Marathi newspaper x= Number of advertisements tobe paced in English newspaper 1 bjective Function: Coverage though Marathi newspaper = 1000 x. and through English newspaper = 1500 %, Total coverage is C= 1000 x; +1500 x2 Objective function i to maximise C = 1000 x, + 1500 x2 2. Constraints: (Advertisement expenditure for Marathi newspaper = 3000 x, ‘Advertisement expenditure for English newspaper = 5000 x2 + Total advertisement expenditure = 3000 x, + 5000 x; Total advertisement budget = € 50,000. Hence, the advertisement expenditure should not exceed the budget + 3000, + 5000 x $50,000 ie 3q45.<50 (@ Maximum numberof advertisements in Marathi newspaper is 10 Mr. Rao, the owner of 2 Reed espe’ of Sserents in two local daily see See sis 1000 people coverage through the advertisement respectively. Each advertisement in dbaily itis 5,000. Mr. Rao has deci ie x10 ‘Simla, minimum number of. ir ao ‘edvertisements in English daily is 6 + WP made of the problem ig Moimise C= 1000 + Sainte NOM *1500%5 (Comrage Maison SPB SSO bet cosmid 410 i x ase (Marathi Newspaper constraint) ate (English daily, constraint) «_Nor-negativty condition) Me peer ns A esearch lab has melts of Cu alloy to make a new alloy. The compostion Mek [__Conpoaion oy] 1 a 1 0 1 1 To make up the new alloy, at least 10 kg of Cu and 6 kg of 25 per kg while melt Il costs ® 30 per kg. Write the LPP for the qi Used to minimise the cost Solution: Let x; - Quantity of melt (in kg) used, x; - Quantity of m 1. Objective Function: Minimise cost C= 25x: + 30%; 2. Constraints: Since, composition of melt is (2 parts Cu + 1 part Ni) 20 20 ae ToT 5 is the proportion of Cu and 5" isthe proportion of Ni in one unit of a Terai ‘melt Similarly in melt I the proportion of Cu and Niisz--z ie.5 and 5 each. eae 2+ Kg of Cun New alloys $x: + x2, Now as atleast 10 kg of Cu is needed, Ors iicay cia icnatindamce (in kg) used. 4x; + 3x22 60 (i) Similarly for Ni, A eae 3. +. LP. model ofthe problem is Minimise C = 25%: + 30% (Cost minimisation) Subjectto 4, + 342260 2x, + 30236 x02 0 (Nen-negativity condition) Seema O_o negatity Condon) __ Problem 3: A manufacturer intends to market a new fertilizer produced from a mixture of 2 ingredients A and B. The composition ofthe ingredients is as follows: Composition Ingredients | Bonemetal | Nitrogen Lime Phosphate A 20% 30% 40% 10% B 40% 15% 40% 5% SS Scanned with CamScannerLUNE rROGRAM ee eae sch thatthe frie” aE "must contain at least 25% Bonemeta, 4 must contain at east 10% Phosphate 4 16/- pet kg for B. 4 ‘of ingredients to be mixed to minimise the The management decision is 1 must be sold in bags of 20k 3. mustcontain at Cost of ingredients is € 20/- Write the LP formulation for material cost. Solution: Let x1 ~ Quantity x2 Quantity of 1, Objective function: 2. Constraints: (Each bag is of 20 kg of ferilizr produced by mixing A and B. x +2 = 20 a ‘A bag must have at east 25% bonemetal. ies 25 , The relevant data ate sivenbelowe Product Py Product P3 €200 7240 245 50 Bhrs. © 2h 10hrs. @ & 2/hr ranetioe 6 hrs @ &2.25/hr oyvri : ae 12hrs @ 2.5/hr fied Ove = 28500 pram 21150 lumber of employees : Number ofhousenpoyehcee me OFFA 2, Dept 8 «15, Dept C= 18 Number of weeks per annum = §9 och department Fora the Gen prcbem a Linear roy fone {ammring Probler, DeCIION scices ecwowscmers Solution: Let x, LUNEAR PROGRAMMING ‘Annial Production of produc P; (in number of unite) ~ Annual Production of product P: (in number of units) Functior selling prices for products P; and Py are € 200 and & 240 respectively les Revenue = 200. + 240 x; = 8x2 xx = 164, Total variable Cost = +. Combined Total . Total Annual Cost = Variable Cost + Fted Cost = 100, + 125; + 285,000 (i), Total Annual Profit P = Total Revenue Total Cost = (200 x; + 240x.)~(100x, + 125% + 285,000) 4 P= 100x, + 15x, - 285,000 - Objective function is profit maximisation, ile, Maximize P = 100. + 115«; 285000 2. Constraints: No.of hrs/employee/week = 40 in each department Since there are 20 emp department A, the weeldy man-hours avaiable in department A = 40 x 20 ‘there are 50 weeks per annurr, “Annual man-hours available with Dept A = 40 x 20. 50 = 40,000 Similarly, Annual man-hours with Dept B = 49 x 15 x 0 = 30,000 and ‘Annual man-hours with Dept. C = 40 x 18x 50 = 36.000 Now, total man hours of Dept A required for producing x units of P: and xe units of (8x, + 10%.) 10125 40,000i.e 4 + 5x: $20,000 for Dept 8 430,000 ie. 51+ 32 $15,000 44; + 12K; 36000 Le. x, + 3x2 $9,000 La ns Scanned with CamScanner(Dept. C constraint) (Non-negatiity condition) ‘of inspectors {and Il who at at e rate of 50 per hour te of 40 per hour with an accuracy of 95%, % 250 per hr. Each time le forthe inspect Er how many grade I and grade Il inspectors to be engaged to to be engaged 1o be engaged paid wages at the rage of & 4.50 per an 8 hrs day. Solution: Let x - number of grade y- number of gt Function: The the Grade I inspectors can check the pieces at the rate of 50 per hour, the ruber of pieces checked by x number of inspectors in 8 hours = $0 x 8 xx = 400x "Non, since the accuracy of inspection is 97% «The expected error = 100-97 = 3% 10.003 + Expected numberof errors by arade [inspectors = 400%%003 = 12x - Sia for rade inspectors, the expected number of errors. = 40x 8xy x05 Thus expected numberof errors in an 8 hou Fo rs day = + Costof eros to the company Sed tnt = Ax(a+ Luge a5 each error costs ¢ 1 Total Cost = = Cost of wages + Cost of errors = G6x-+ 207) + (12K + 16)) = aK + 36 + The objective function sto minimize C= 4 ¢ 36y 16 DECISION scRNCes memes es rca 2. Constraints: (0. Number of pieces checked in 8 hours = Pieces checked by Grade I inspectors + pieces checked by Grade inspectors = 50XBxx4 4x8 xy = 400K + 320y "Now, as atleast 2000 pieces ae tobe inspected = 400 x+ 320y 22000 le. Sx+4y225 (i) As 10 grade inspectors are avaliable withthe company xsi0 Similarly, y <5 3. Hence the LP. model forthe problem is: C= 4x + 36y (Cost minimization) 0 Sx+4y225 (Output constraint) xs10 yss, xy 50 (Non - negativity condition) Problem 9: A firm produces 3 products A 8 and C. It uses 2 raw materials land 11 of 1n be used for production of A, B and C. Product A requires Lnits of raw material II per unit. Corresponding requirement ‘and 3 units of raw material I and Il respectively and per unit of C, 5 units of The labour time to produce 1 unit of A is twice required to produce 1 unit of B an times requited to produce unit of C. The entire lat the firm can produce equivalent of 3000 units of product A. The minimum ‘demand for the 3 products is 600, 650 and 500 units respectively. Assuming profits per unit of A, B and C are & 50, & 60 and € 80 respectively, formulate the LPP. to maximise proft satisfying constraints. (P.U. MBA - Dec. 05) Solution: Let x y and z be the production quantity of the products A Band C respectively, : 1. Objective Function: Since the per unit profit of A B and C is € $0,€ 60 and f 80 respectively, ‘2 Total Profit = 50x + 60y + 802 ‘Objective Function is Maximize, P = 50x + 60y + 80z a in asl Scanned with CamScannerLINEAR PROGRAMNang pesos, 2. Constraints: {paw materia: Since, eac uit ofA B and C reaures 3,4 and 5 units of ray And as a 5000 units of 3x + dy + 52 5000 Similarly, Total numberof units of raw mat [And as a 7500 units are avaliable, hence we must have 5x + 3y + 5257500 (i Labour Time: Let a,b and ¢ be the labour time required to produce a unit product A, B and C respectively. Now, asthe labour time required to produce 1 unit of Ais twice that for B and thrice that for C a= 2b and a= 3 ined = 5x + 3y + Sz 1 1 = b= Laandc=t Saandeste Total labour time required for production Faxxtbxytexz saxstaela Bris: p Now, the labour force of the company can produce equivalent to 3000 units Product A "| 5 Total labour time available = 3000 xa = We must have axelaysl 3+ Fez S 30008 eee jex+dyed 3+ Zz $3000 ie. 6 + 3y + 22 < 18000 212600, y2650 and 2 500 DeIION scuces secon ecenees ean rans, 3. Hence, the LPP. formulation is Maximize P = S0x + 60y + 902 subjectto 3x4 4y + 52 «5000 5K + 3y + 5257500 6x + 3y + 22 18000 x2 600 y2 650 22 500 ny 220 Problem 10: A refinery makes 3 grade Crude ol Fcan be used in any grade but th Grade | Selling Price per (Profit maxim must satisfy the fo ‘Specifications litre @ A 748 "Not less than 50% crude D Not more than 25% crude € 8 50 (Not less than 25% crude | Not more than 50% crude E c 49 No specications ‘There are capacity limitations on the amounts of 3 crude elements that canbe used rude | Capacity (kt) Price per Litre @ Ds. 500 495 E 500) a5 F 360 485 Formulate LPP to maximise profit. (PU. MBA ~ May 07) Solution: Here, the specifications of A, B, C are expressed in terms of minimum or maximum contents of D and E. Also for calculating the profit we have to take into account the cost of crude oils D, E, F as well as selling prices of A 8, C. Hence, it is advisable to consider the decision variables as follows: let 240 ~ Quantity of D (in litres) in petrol A; xsx - Quantity of (nlite) in petrol A ‘4g — Quantity of F (nites) in petrol A; xa0 ~ Quantity ofD (in litres) in petrol 8 “xee Quantity of Ein tres) in petrol B_ >ar— Quantity of (nites) in petrol B cp Quantity ofD (in litres) in petrol C ; xee~ Quantity of En tres) in petrol € ee = Quantity of F (in litres) in petrol C ee oa. Scanned with CamScannerNEAR PROGRAMMING ee 1. Objective Function: The total production quantity in litres) of petrol A = x10 * Xue 1% litres) of Band C are and (co + Xx +%c) respectively. Now a the seling prices (pe Band Care & 48, $0 and 49. 39 petrols) = 48 (x10 + xar +X) + 50 (Heo + Oe iar) + the crude oll D used = xp + X80 +%0 the crude oils E and F used are (xa: + Xe +e) and 1495,€475 and 8485 Total Cost (of crude oils) = 49.5 (Kap + Xe0 +o) + 47.5 (kas + Kae +X) + 485 (ur + Her +40) _. Total profit?) = Total Revenue - Total Cost to + Hat Hue) + 50 (860 + ex tHe) + 49 Okco + Xe Ha) — 495, Yeo +80) 475 Gear + Yar He) ABS (har + Yer x0) The objective function sto maximize P2~15i0 + O5kse~O Siw + 05 r40 + 25% 2. Constraints: (© Capacity: Since the available quantity of Dis $00k «ie. 500 10° litres 2 We must have, of D used < 500% 10? + kco $500 x 10° crude oll se + Xe + He 500 x 10? rude oll, xa + Xr + XS 360 X 10? (0 Specteations For petrol the quantity of crude oil Din iti. ap should be at least 50% (by volume) ie rao 2 050 x (Quantity of petrol A) le m0 205 (tao + Xue + Xu) ie.05%0-05 xe -05 520 emp ~ Xu = Xu 20 Also the quantity of crud ie. me 50.25 x Quantity ie rg $0.25 fo + Ha + -O5tep + L5ice + 05 Xr 12. Xag should not be more than 25% Mae + ter 0 fol Chas no 3. Thus, the LPP is, Maximise and ¥9¢ $05 (ego + Ho + He) = 0500 - 05x $01. xa ~ Hap — 144 $0 changing sign P15 rp + 05xu-OStay + O5ta0 +2510 + LStqe— 05 keg + L5kee + 05 xcp Subject to: M40. Kat Ka YY Ma, Keo Kees Her 20 (profit maximisation) (Capacity constraint for >) (Capacity constraint for E) (Specification of & for & (Specification of 8 for D) (Specification of 8 for (Non - negativity condition) Problem 11: The manager of an oil refinery must decide on an optimal mix of two Possible blending processes of which the input and output per production run are given below: Process Input (Units) ‘Output (Units) GrudeA Crude B | Gasoline X Gasoline ¥ 5 3 3 8 4 5 4 4 ‘rude A and B is 200 units and 150 units respectively. Market requirements st 100 units of Gasoline X and 80 units of Gasoline ¥ must be produced. The profit per production run from process I and process ll are ® 300 and & 400 respectively. Formulate the above LPP problem. (P.U. MBA - Dec. 09) Solution: Let x - Number of production runs of process 1 y- Number of production runs of process 1, Objective function: Since the profits per production run from processes I and Il are Sore - soem haat, ie mo~ ku : € Kap ~ Bag + X20 v= by changing the signs a Hist vaca cn ia Scanned with CamScannervex moceavnans Becton scxevcg§ 2. Constraints: {@ Crude A: Each production run of process I requires 5 units of crude A as input wave 5x + 4y 5 200. 3x + Sy and as the availability is X Sales: Each production run of process I produces 5 units of Gasoline X ‘output. Similarly, each production run of process Il produces 4 units of X. Hence, total production of Gasoline (through x and y number of respective Sx 4y 2 100, (iv) Gasoline Y sales: Similar market requirements le ry 220, 3. Hence, the LPP formulation: Maximise P = 300x + 400y Subjectto 5x + 4y < 200 3x4 Sy < 150 Sx 4y 2 100 {and the number of man-hours available with the firm per week are Sebo pee ail aciecaes i alge 0 5 een $ Fats Copa pt Profit on sale of A is © 30 per unit as com ‘A x: - No. of units of B produced per week. Max P= 30%; + 20x, subject to 10%, + 6x2 < 1000; 5x, + 4x0 $600; x, 32 0) + BS 100 ts F600 Fu 322 ne Winner Sports wishes to determine the number of advertisements to be Selected monthly magazines A, 8 and C. His objective isto advertise tal exposure to the principal buyers of the expensive sports entage of re is the number rents placed multiplied by the 5. The following data may be used: ‘number of principal by Magazines A & rl Readers lakh 06 lakh O4 lakh Principal Buyers 15% 15% ™ | 5,000 4500 4250 int at most is ® 1 lakh for the advertisements The owner has already ine A will have no more than 6 advertisements and that 8 and C model for the problem, fo of advertisements to be placed in A,B, C respectively. Max E = 15000 x4 + 900% + 2800 x month. They use these materials to make three basic pharma products viz. $-10-5, 5-5-10 and 20-5-10 where the numbers in each case represent the percentage by weight of A.B, C respectively in each of the products. The cost of these raw ‘materials are & 80, € 20 and € 50 per ka. respectively while the inert ingredient costs 20 per kg. The selling price of the three products are 40.5, € 43 and & 45 per kg respectively. There is a capacity restriction of the company for product 5-10-5 so that it can not produce more than 30 kg per month. Determine how much of each of the products the company should produce in order to maximise its monthly profit? (Ans: x, x2 % ~ Quantities produced per month (in ka) $10, 20-5-10 respectively. Max. P = 16x + 17k + 10x Subject to $30; Su + Su + 208 100000; 105; + Sx: + 5318000; 5x, + 10x: + 10% £12000" x, 01020) ‘A company is making two preducts A and B. The cost of producing one unit of product A and B is & 60 and € 80 respectively As per the agreement the company has to supply atleast 200 units of product 8 to its regular customers. One unit of product requires one machine hour whereas product B has machine hou available ‘abundantly within the compary. Total machine hours availabe for product A are 400 hours, One unit of each product A and B requires one labour hour each and total eaieeananens irs enn or 1a Scanned with CamScannerLUNEAR PROGRAMMING a DECISION sciences 500 labour hours are available. The company wants to a if a of Pecmowscwncs nen cain, reduction by satiying te given requrement. Format the over & ao — a (ana y= No.of units of A and B to be manufactured respectively | Sa oana cc Gx; Subject to xe 200; A007) mi +X S800, x, 2220) ee ve A city hospital has the following ‘minimal daily requirement of nurses. in : t 5 | ta & Period 1 2 3 4 5 6 : sf 2 fats Clock time | 6:10am | 10-2pm | 2-6pm | 610pm | 10-2am | 2-6 am z 5 4 2 [af Min no. of[ 2 7 6 8 20 6 Available Hours 160 20 ne one Formulate the LPP. (P.U. MBA - May 06) (Ans. Max P = 4x + 8y + 67 sti x + By + 22 < 160, 3x + dy + 22 < 120, ‘The nurses report to the hospital at the beginning of each period and work for 8 hours consecutively. The hos nts to determine the minimum number of nurses to be employed so that there will be sufficient number of them available for ‘each petiod. Formulate the LP model for the problem. (Ans. X, ta» Xe No. of nurses reporting during periods 1, 2, Min, N= 4X2 4%) 4 Xe 45 +m Subject to x1 +027; +4215) Ma tH5220; % 4% 26 H+ Hy Ho X20) ‘A-company produces two special types of soaps X and Y for which the following data , 6 respectively. Xs +428; is available. Per Unit x Y Selling price ts 2 Direct Material: A 2units @ & 2/unit 3 units @ ® 2/unit 8 Lunit @ € 4/unit units @ & 4/unit c - unit @ & 1/unit Direct Labour man hour @@25/hr | 1 man hr @& 25/hr Variable Overheads a qs The fixed overheads are © 1500 per month. The quantities of materials AB, C avalable for production are 500, 400 and 200 respectively per month. There are 2 workers who work for 8 hours a day for 25 days in a month. The per month market ‘demand for X and Vis at least 200 and 150 units respectively. Formulate this as a LPP (Ans: xy - No. of units of X and Y produced monthly. Max P = 6.5x + 6y - 1500 Subies 24+ 375500 x + 275400) y S20 x ys400 x 2200, y2150 xy ‘A manufacturing company makes 3 products, each of which part of manufacturing process. The company can sell Inula but its production cpabies a ire 3 operations as a the products it can related data are as Eee SEN nine WH ia oe 128 De +y +225 80;x 9,2 20) ‘A company machines and drills two castings X and ¥. The time required to machine and drill one casting including machine set up time is as Casting Machine Hours Driling Hours x 4 2 y 2 5 ‘There are two lathe and three drilling machines. The working week is of 40 hours there is no overtime and lost time. Variable costs for both castings are & 120 per unit while total fixed costs amount to € 1000 per week. The seling price of casting X is nd that of ¥ is € 360 per unit. There are no limitations on the number ings that can be sold, The company wishes to maximize profs Formulate programming model for the problem. (Ans: X-casting X y-casting Y per week Max. P = 180x + 240 y - 1000; St. 2x+ y's 40, 2x + Sy $120;xy 20) {A firm produces three products, These products are processed on three different ‘machines, The time required to manufacture one unit of each of the three products dnd daily capacty ofthe three machines are given inthe table below: Machine ime per unit (minutes) Machine Capacty Produta | product2 | Product3 | (Minutes pet day) ™ 2 3 2 “40 Ms 4 = 3 470 My 2 5 ef 420. bx i i ‘er This required to determine the daily number of units to be manufacture for each Wp por unt proaic Zand ist At 3nd repel Nt Sed are consumed in the market. Form assured ta a he amounts produc ea (ans: Max P= 4x, + 3X2 = Product 1% ~ Product 2,35 - Product 3 dally produced: (ped subject to 2 + 3+ 2 5 A40 fu + 30 $ 470, De + StS ONY 2 0) as Scanned with CamScannerLINEAR PROGRAM [DECSION SCIENCES The profit per unit of is required to supply 209 . ‘of Mi is requ yan tors tasencugh ours avalble to manufacture any of units of product B. M; has 400 hours available. Product A and B both one labour hour each and company has 500 labour hours available. To ine the number of units of A and B to be manufactured, sat conditions. Formulate the LP. (P.U.MBA-Nov.03) (ans: x- A produced, y- B produced; max.P = 60x + 80y, subject tox = 200, x < 409, x+y s500;xy20) PART Il - GRPAHICAL METHOD OF SOLUTION, ‘After a Linear Programming Problem is expressed in the LP format, the next step is to jefore going for the solution (ie, the set of values of the deci ing points must be noted: lem can have many feasible solutio ‘Solution: This is @ basic feasible solution which optimises ‘The two decision variables say x: and x2 are plotted along the two coordinate axes X ar ‘on a graph paper. Only first quadrant of XY plane must be considered as x; x2 2 0. ‘Steps for obtaining Graphical Solution: Step 1: Formulate the given problem in LP format Step 2: Plot the graph constraints: all the terms. Change the inequality from2tosorstozinit (i Replace the inequality sign (< or 2) in a constraint Si rth ‘8 constraint equation which represents a stra mm ‘an equality sign (=). This gives (i) Plt the constraint in: For ploting the line, identify two points on the 2s Theos convenient pots te ipo ‘ais (obtained by putting x; = On the constraint equation and solving it 126 Soe | Lien Prosaaseanes Y axis (obtained by puttin by Putting x: = 0 in the equation and solving it for %,). Note that the lines xy = a and x = b from the xi tar 5 parallel to Y and X avis at distances @ and b line divides the XY plane into tw. into two parts - one which includes the Origin and the other away from the origin, cau (id) For constraints of the type cor line. For cor about the const f only. Use arrows type (=) the region isthe line Je regions. Mark the region whichis common to a hich could be an open or closed polygon Step 4: To find Optimum Solution: Ss _{A) Corner Point Method: ~ (i) Identify all the comer points of the feasible region. To obtain the equations ofthe corresponding (intersecting) constraint necessary. 1S regions. This isthe Featible Region, ir co-ordinates solve imultaneously, if Find the value of the objective function (say Z) at all these corner points. For @ maximisation problem, the co-ordinates of the comer point where Z is maximum give the solution. Si imisation problem the co-ordinates of the comer point corresponding to minimum value of Z give the The corresponding value of Zs its optimum Value Zar Of Zn 3) 150-Profit or Iso-Cost Line Method: Alternately we can use the Iso-Profit or Iso-Cost Line Method: (Draw an objective function line (passing through the feasibie regio) lke @ constraint 1 by considering some postive value for 2 arbitrarily. This is an io-profit(so-cos) as all points nit give same value of Z If the problem is of maximisation (minimisation), go on shifting ths line paralel to itself away (closer) from (to) the origin, until further movement takes the line ‘completely outside the feasible region. iy The co-ordinates of the point touched last by the line give the $2 sand xg and the value of Z fr this point gives valve of Zou (Zon) 17 lution ie. values of Scanned with CamScannerDECISION scENces nec (i) Consider the second con: straint 2x + wegettheconsvant equation ag,” ** Sea neu byequaty gn DECISION SCHNCES ‘tustration 1: Solve the following problem graphically = x4) Maximise Z = 3x+4Y jon as, Subject to xeys6 d+y=8 aeys8 Put x=0 Fi xy20 X and Y axes. en) hi along the X a ‘Solution: The decision variables x and y are plotted slong “GE, 8 fs point on ¥ ais Puty=0 228 xe4 *-__D(4, 0) is point on X axis and «. line CD is the constraint line (2) 2. To find the Feasible Region: Dom eo. is x + y$6 ie of the type <. Hence, matk the region towards (1) by arrows as shown above, isthe feasible region as marked in points ODEA (iv) To find the co-ordinates of point E, which i a point of intersection of lines (2) and (2), solve their equations simultaneously. Thus, we have xty=6 ® ary 8 @ Subtracting we get -x+0=-2. 2x=2 2 From(l),y = 6-x=6-2=4 Co-ordinates of E are (2 4) 3. Find the values of Objective Function: Finding values of Z = 3x + 4y at the comer points ©, D, E and A by using their co- ordinates as at (0,0) Zo = 30) +4 =0 D (4,0) Zo = 314) + 4(0)=12 +46) =6 +16 = 22 +4(6) = 24 is a problem of maximisation. Hence, the optimum value of 1g the above ie. 24, This coresponds to point AO, 6) ¥ = 6s the solution for which Zmac = 24, sm of minimisation ie. Minimise z = 3x + 4y, the y= Owher® Zrn = 0. 1 equation of the frst constrain i ses B/SuER eR I line. Its intersection with Y » O+y=6ie ya6 +N yi. A(0, 6s the point on Y axis Similarly, putting y = Oin @) we getx= B60) is the point on X axis, Jong tere points A and 8 we Gt the consti ine 1 Se Note: If this would have been a probler Scanned with CamScannerLINEAR PROGRAMMING fees mustration 2: Minimise Z= 6x1 + 5 x= 10 3) lel to Y axis ata distance 10 from origin ie. a vertical DECSION scinces ppeiaeat nar crane 2. Tofind the Feasible Region: ) Is of type 2... Mark the region away from origin abi by Mark the region away from origin about also of type > ‘extends upto infinity. () To find co-ordinates of F, we solve equations of intersecting lines (1) an simultaneously. (3) gives and Subtracting, «From (0) x2 # 10-44 = 19-4(3) =10-6=4 44 . To find values of Objective Function: Z = 6x; + Sx: at comer points: at Al, 10), Ze = 6(0) + 5(10) = 50 ar(3.4) 2 = )+5)=29 ao(i.o} De (3) +0=45 at E10, 0), 2 = 6(10) + 0= 60 4. To find Optimum Solution: This is a problem of minimisation. The minimum value of Z occurs at the comer point G Thus the solution sx: = 3,10 = 40nd Zn = 23. ‘Note: If this would have been a problem of maximisation ie. Maximise Z = 6x; + Sx> the solution would have been Zn, = =, corresponding to point (10, ~) as the region extends upto =. Hence the solution would have been unbounded. i Scanned with CamScanneror genera nes Programing robe hawng nurunowns (econ Vadalag xX — Yqand m constraints wherem < : 1 ase Solution: A soon obtained by seting (wm) number of varies Grom 0 enables Xs Xo x) to0 and thus solving the m equations in m remaining variableg ‘salle a basic solution The variables (n-m in number that are set to zero are called fables and the remaining m variables are called as basic variables ‘which form the basic solution. a (i) Optimal Feasible Solution: Any basic feasible solution which optimises the obj function is the optimal feasible solution (i) Degenerate Solution: A basic solution for the system of equations is called as degenerate solution if on become equel to zero, Problem 1: Minimise 2 = 80x; + 120%. Subjectto +059 m22 23 20%; + 50% < 300 my 620 Solution: 1. To plot constraint ines: © % +559 + Constraint equation is x, +2 = 9 p= ued = AS) mr ame 5 BOO) Constraint equation is x = 2 ie. Une paalel to axis at (2.0) 23 + Constraint equation is, xse 3 lie Line parallel tox ais at (0.3) (9) 20% + 50x < 300 “+ Constraint equations 20x; + 50%: = 300 le 214 5q230 Points are C(0, 6) and DAS, 0) 1a Decision sciences LINEAR PROGRAMMING Mark region towards origin. =) Matk region away from origin. =. Matk region away from origi. s. Mark region towards origin. le region is the polygon EFGHE, common to all the above regions. fon (2) anc F lies on (3) and (2) 2 ms3 o/ From (Wn +3=9 226 «FG 3) G lies on (1) and (4) Solve (1) and (4) simultaneously mt = Mate = a+ 5 = 30 0-3x = -12 2 xe wepes (1) one 62) Subtracting “a From), =9-x = 9-425. +. (5,4) = From), 14 =2 1s Scanned with CamScannerLINEAR PROGRANNING pean il Zin = 520 at E, = 2.x, = 3is the optimum solution for which Zn = 520. Problem 2: (Multiple Solutions) Maximise Z= 3q+5m Subject to 123 + 20%; < 40 nnd <4 mm20 Solution: 1. Constraint Lines: 0 12q + 20% = 40 le 3 +54 = 10 Puty=0, .54=10 A, 2) =0 3x =10 ® DECISION sciences nn _UNEAA PROGRAMMING . le regio 3. Values of Z = 3x: + 5x2 at corner points: at 0(0, 0), Zy=0+0=0 mE a2) sm 0 hhas equation 3x, + 5x, = 10 i.e. 5x3 = 10-34) 3x + Sr = Z Le. the objective function + 5x). Therefore they have the same slope. Hence they are parallel, Thus, sation problem, the farthest iso-proft fine coincides with line AB and hence, ‘co-ordinates of all points on AB give the optimum solution values of x, and x: and corresponding value of Z i.e. 10 gives Z max. Thus the problem has multiple solutions, coresponding valve of ZLe10 vei 2net S ee———— 15 Scanned with CamScannerLUNEAR PROGRAMMING a Problem 3: Solve using Graphical method Minimise Z=n+4n Subjectto Su 2% noes utes 9 mm 20 Solution: 1L. The Constraint Lines are: (The constraint is 5x: 2x2 i. 5x: -X22 0 Const Putx,=0 om #028 £0 +.0(0.0) (if we put x; = 0 we get same point O(0, Hence to get another point online (1) put for x: some value other than 0 say 1% = 2 (Note this step). =O 3 y= 10 2 Points A2, 10) i) mates 5 - Points 80, 5) C15, 0) (ip Constraint is 3x, + 30 = 9 which sof type Hence, the const ie Inte + Points (0,9), £(, 0) 2. To find Feasible Region: (The frst constraint is of type ‘a. But the line Hence, consid (Note this step) : (2) passes through origin itself. 27 test point lying in the right portion about the line, say (1, 0) Raine ae pen DfcION scmnces LINEAR PROGRAMMING Hence from (1) we have, test point say (0,2) ying in the left portion about the line (1) LHS = 5(0)-2 = ~2 which is < Oe. it violets the constraint. Hence as the constraint is of type '2, the region will consist of the line (1) (for =) {and the part on right hand side ofthe line (1) (for >), Gi) Since constraint (3) is of type ‘=. The region here wll consist of points lying on the line (3) ony. (iv) The common region to above three regions isthe feasible feasible region regions common to constrains (1) and (2) ie. it isthe segment FE. Co-ordinates of Fare obtained by from (2) from (3) Subtracting -24 = -4 4 we gee w= S-m=5-2=3 To fed vali of Z= +4) wt he corer poles: The feasible region tne segment FE “at F(2, 3), Zp = 2+ 4(3) = 14 at E(3, 0), Ze=3+0=3 4. Optimum Solution: ‘We have Zmin = 3 at E(3, 0) Problem 4: (Unbounded Solution) Maximise Ze as 5y Subjectto x+y 21 ae-yz-1 Ax-2y $1 aaa ea = Scanned with CamScannerLUNEAR PROGRAMMING easton cENcES Solution: 1. Constraint Lines: () @ xtyet 8a, 0) eryend 's0 as to make the rhs. +ve, we rewrite it as = 2x + y<2. pois ce», 0(-3.0) weave 8 mike) DEON scunces Sc near rroceanans 2. Feasible 2 type &: Mark region away from origin *. Mark region towards origin, ) Constraint Type 2 Mark region towards origin i shown in Fig. 1.6 which extends upto infinity. a) x4y= 1.0) x2 ee ® Serer : sgbmitigeicnna-badt x(t 3) 3._ Since the problem is of maximisation, the point in the region lying at infinity say (© ©) will give Znas = = which is the maximum value. Hence, the soltion is unbounded. Problem 5: (infeasible solution) Minimise Z=1OK+% Subjectto +205 4 mite 2 $ une 0 Solution: 1. Constraint Lines: @ +2024 ~() Points (0,2), B(4 0) Gm +0=5 ~@ + Points C(0, 5), D(5, 0) 2. Feasible Region: (0 ()isoftype's —_«. Region towards origin. isoftype'>’ «Region away from origin. 139, Scanned with CamScannerLINEAR PROGRAMMING en eaet Fig.1.7 ro region common to the above two regions (in the first quadrant) hence there is no feasible region and hence the problem has infeasible soluti Problem 6: Upon completing the construction of his house, Mr. Sharma discovers that ine scrap are in usable form of plywood and 8 sq, feet of white jood and 16 sq.feet of white pine are required to finished products to a local furniture store, Mr. Sharma ‘and & 100 on each book case. How can he make fofitably? Formulate the LPP and solve using graphical Per Unit Requirement (sq. feet) Plywood White Pine | Profit per unit Table 16 8 25 Book case 2 16 100 Availabilty 100 80 - Tet x; - Number of tables constructed and 5 de inter Asoka » Function: Total Profit. Z= 125 x, + mi 2 ve fa 125 x, + 100%; to be maximised. (Plywood constrain is Pymcod Wt Plywood required is (16x + 12%) and 100 sq, feet are 2.16% +124 $100 ie 4 + 34525 (i) White Pine constraint: White pir wine Pine required is (8% + 16x,) and 80 sq feet are 284 +1680 en +2010 a st 140) Laven PROGRAMMING a 833. A(0,833) @ ‘of type 's._». Mark regions towards origin the feasible region is OBECO. a 4 ned foe, nago-ges towns a E43) Scanned with CamScannerDECSION scunces a _LUNEAR PROGRAMMING, at corner points: +100 x) at corner pol 3. Thus, the LP Model for the problem is 6, Value of @ = 125%: snmsas oe 2,0} 25=22() +100 = 702 sect agon ta at E(, 3), Ze = 125 (4) + 100 (3) = 500 + 300 = 800 IK + Sq 2 44 atC(0,5), Ze =0 + 10045) = 500 7. Optimum Solution: We have Z max. = 800 at E(4 3) (Non-negativty condition) Hence. Mr. Sharma can manufacture x: = 4 tables and x2 = 3. bookcases to eam * : Problem 7: Mohan Breweries this wo betting ants cH ie ner ae 7 and the cther at Mohan-Nagar. Each plant produces three drinks, whisky, a 2 ) juices named A.B and C respectively. The numberof bottles produced per day are as foll Points are C(0, 40), D\"3 , 0) ie. (13.33, 0) Plant at 2a +5 = 44 a a meee SES Naeide\iaa‘aaiga Whisky A 1500 1500 Beers 3000 31000 Fruit Juices € 2000 $000 [A market survey indicates that during the month of April there will be @ demand 20000 bottles of whisky, 40000 bottles of beer and 44,000 bottles of fruit juices. T ‘operating costs per day forthe plants at Solan and Mohan-nagar are 600 and 400 monetar Uits For how many days each plant be run in April so as to minimise the production c hile sil meting the market demand? Formulate the model and provide graphical solutio Solution: Let »; - No.of days in @ month for which pant Sis run 2 No. of days in'a month for which plant M is run, 41. Objective Function is to minimise the monthly total production cost which Z = 600% + 400% 2 Constraints: (). Matket demand for whisky is 2,000 bottles 1500; + 1500% 2 20000 ie 34+ 3240 ) Market demand for beer is 40,000 bottles 3000; + 1000.2 40,000 ie 3x; +m 240 (i) Market demand for frit juices is 4,000 bottles * 2000 5 + 5000 x; 000s + 5000 x2 44000 leu + Se2aq 1a Scanned with CamScannerLunean PROGRAMING SSeS a yg 5. Feasible Region: Since all the constraints are of the tyPe "= ‘about these lines. Thus, the common regi rais-CGF - X® nce, mark the regions away from origin he feasible region, Which is the open EaIION ScRNCES Lean rRoceannans 2. Constraints: Total loan amount is ® 21,00,000 maximum x+y $21,00000 Amount of personal loans (x) can not exceed 40% of car loans (y) Scanned with CamScanner esandert as consi (2) implies constant (1) aso) ta Sn etn eas 2) and) maton a ters oie uossioy Bx +x = 40 2x5 gives ie 100x-40y< 0 153+ Sia = 200and 3. Thus, the LP formulation forthe problem is as + Ssstal Oh @ Maximise Z = Olly + 0116 Prof maximisation) Subtracting Tx, = 156 Subjectto ——x#y $ 200000 Fund Avalabilty constrain) 100x-40y <0 (Proportion Constraint) te n= en xy20 (Non-negatvty condtion) ce From (2) x2 = 40-34; = 40-3 (12) = 4+ G(12 4) 4. Graphical Solution: The Constraint lines are 0 x+y = 2100,000 ® 6. Value of Z = 600 x; + 400 x2 at corner points Points A(0.2100000), B(2100000, 0) Zc = 0+ 400 (40) = 16000 @ 100x-40y = 0 @ 40044) = 7200 + 1600 = 8800 ie toy = 100xiey= Ay iey2254 5 a= Se cn See Pom i yO) 2°80 arity = E000 Gy 35 t0 Keep the production cost sex 2200100. sonq50 «120000, 20000) Problem 8 A smill bank is allocating mavimum of €23,00000/- for personal and car terest rates per annum are 1195 for cer loans and 13% for personal loans. The loans are repaid at the end of a one-year period. The amount of personal loans can not exceed 40% ofthe ca loans. Past experience has shown that bad debts amount to 1.2% of personal loans. Solve the above problem to find the optimum loan allocations and banks profit. Solution: Let_x- Amount o be allocated for personal loans. y-- Amount tobe allocated for car loans. 1. Objective Function: The loans are repaid at the end of 2 year with 11% and 13% {or car loans and personal loans respectively The bad debts are 1.2 of ll Ly, 28 py (22) ys, = Olly + (0.130.012 = Olly + 0.118 x which is to be maximised EEE taUNGAR PROGRAMMING a a ae 5. To find Feasible Region: LHS = 100(5)-0 1 500 wtichis> 0 oe go (25) = 200 whichis >O Similarly consider (0, Hence, this part viletes co "Now, consider points in other Ls = 0-40) Also at (~5, 0, LHS = 100 Hence, this part satisfies cons line 2) (i) The common region is the feasible region ie. the region ODAO. To obtain D, solve (1) and 2) simultaneously, == 200 which is <0. +0=~ S00 whichis < 0 2), Hence matk tis left hand portion about 5 ya A resi og tse nena (mea : 4 le (232%) x= 2xono00 X= gp (2.00000) = 6,00,000 = From (1, y= 21,00,000 x = 2,000 ~ 600.000 = 15,00,000 To find values of Z = O.1y + 0.18% Z=0+0=0 ie. 0(600000, 1500000), 0:1 (25,00000) + 0218 (00.000) = 165,000 + 7000 = 235,800 at, 21) ie. A(, 2.00000) 2= 011 21,00000) + 0 = 2 31,000 Optimum Solution: Thus, we have Zee ® 235,800 at 0(600000, 1500000) Thus, the optimum loan allocations will be & £15,00,000 as car loans which gives a profit of: Problem 8: Solve graphically the folowing Linear P Minimize Z= 3X, + 5k Subjectto =3%, + 4X5 12,2K1-% 2-2 DK + 3X2 12.154 Xo2 Zand X20 Shee oe 500.000 as personal loans and 35,800 tothe bank. DECISION SCHNCES, Lavean pRoGRAdNen Solution: 1. The Constraint Lines are () = 3K + 4X) = 12 ® PutXi=0 aX =12 X23 00, 3) PutXs=0 2. -3X.= 12 %=-4 B-4,0) i) %-Kz -2 Since, the ths is - ve we change the sign through out and also change the inequality to 2K, + Xa $2. (Note this step) @ @ (5) ° t ie. Fig. 1.41 a Scanned with CamScannerA LUNEAR PROGRAM peasouscmners « (, (rocited) and (4) * mi oe te the regions away fom ogi 2. Feasible Region: The constai Mark regions containing o” Which are of the type Hence the feasible rei From (5), X= 2 : 2K; + 302) = 12 & Mhew-6=6 m3 13.2) is intersection of (4) and (5) tersection of (1) and (4) From (), X= 4 + From (1),~3(4) + 4X, = 12 4X; 124122 24 2 eb + KC 6) Lis intersection of (1) and (2) From (2),~ 2Ki + Xo = 2 nn 04) ie 8K = 8 = 3K +O = 12 fl Subtracting, - 8%, =—4 4 sremay Hem 2026) +8 eatson saences LNoAR PROGRAMMING ‘Misintersection of 2) and (@) == 2 a Wi +3% = 12 o ‘Adding, OX = 14 uz we Bet fag ae ae 3) 3 med ag Zi, = 34) + 5(2) = 22 Ze = 3(4) + 5(6) = 42, x06) 5) = Baw 12490 102 2a EES = 204 (3) . (2) 9,38 2a {29 947079 © oooh = 1975 4. Optimum Solution: We have Zn = 19 at The optimum solution is X= 3, Xp = 2 where Znin = 19. 1a) Scanned with CamScannerLINEAR PROGRAMMING someones tion plan of medicines is preparing a produ vnxo which either of medicines can be put material to fil 1000 bottles of A and one sole for this operation. The profit is & 8 ve LPP. -Atobe produced Btobe produced. er of medicines kes 3 hours to prepare enough 1, Objective Function: Total profits P= 9X; + 7Ke The objective function isto Maximise P= 8X; + 7K 2. Constraints: (Ingredient constraint: The avilable ingredients can be used to produce 20,000 bottles nd 40,000 bottles of B at the most. Hence, X, $ 20,000 and $49,000. (i Bottles constraint: There are only 45,000 botties available for filling up the medicines A and 8. «+ (Total bottles = Xi + X:) $45,000 (iy Time constraint: 1000 bottles of A require 3 hours for preparing the_fillup rata Hee tiered rene of Ai Hee ote of 2% atnees (2% ned hours Silay, Xs bots of Bw ned (22) hours ewe x ++ Total time requited = pep + 798 an snce the te aelable is 66 hours we % rmusthave, 2 6 ie 610 3K + 10566000 Maximise P= 8X + 7X; Subjectto x: $20,000 Xe 40,000 (Profit maximisation) Angredient Constraint for A) Angredient Constraint for 8) X1+X2545,000 ottles constraint) 3K: + X $66,000. (Time Constrang) Xi %20 (Non-negativty condition) iT rar eaison sexences LUNeAR PROGRAMMING Graphical solution: Scale used is 1 unit: 10,000 4, Constraint Lines: oY axis through A (20000, 0) ~@ 0X anis through 8 (0, 40000) Xp + Xp = 45,000 ® Points: C (0,45000) and 0(45000,0) (iv) 3X + Xe = 66000 ® «+ Points: E(0,66000) and F(22000,0) Region: Since all the constraints are of the type '<, hence mark the region rds origin about all the constraint lines. The common region is the feasible ‘region OAGHIBO. is Scanned with CamScannerLUNEAR PROGRAM DECI cies Point G is the intersection of lines (2) and (4) From (3), X, = 20000 2. rom (4), 3(20000) + Xe = 66000 cq = 66000 ~ 60000 = 6000 6 (20000, 6000) Point His intersection of 3) and (4) we get, @Xi- Xi) + (K-%) = 66000~ 45000 ie 2X = 21000 21000 2 x, = 22%. 10500 and +. from (3), Xp = 45000 X = 45000 ~ 10500 = 34500, (40500, 34500) Point Ii the intersection of (2) and (3, From (2), Xp = 40,000 From@, 1 = 45000-X + 45000 40000 = 5000 (5000, 40000) 3. Value of P = BX; + 7Xzat the comer points: 000) = 160000 + 42000 = 2,02,000 (34500) = 84000 + 241500 = 3,25,500 +7140000) = 40000 + 280000 = 3,20,000, aie ce Rar Heerpectad prom IZS OO. ys Problem 41: A company mar daily. The miture consis of two ingredient F costs 8 pe late the LPP and find the quality of (P.U, MBA - Dec. 10) each ingredient used to minimise cost 1s Decision scunces LINEAR PROGRAMMING 1. Objective Funeti 5 per kg and Fa costs € 8 per kg. Hence, the the mixture is (5x * 8y) which isto be minimised, Hence, the objective function is to rminieise, Cm Sx + By 2. Constraints: () Mirture Quantity: As the mixture must be 500 kg, we must have, x+y = 500 (i) F, and Fz requirement: The feed should not contain more than 400 kg of Fi. Hence, we must have xs 400 ‘Similarly, it must have minimum 200 kgs of F : y 2 200 3. LPP Formulation: Thus, the LPP formulation is Minimise, C= Sx+ ay Subjectto x+y = 500 xs 400 y 2 200 ie. line parallel to Y-axis through C(400, 0) y = 200 8) 0.200] t @ (800, 200), cI (0) (400.0) (800,9) Fig. 2.33 153 Scanned with CamScannerUMEAR PROGRAM, DECISION ScHENCES () Feasible Region: the point is ofthe type =! Hence the region wll consists of the points lying oy ofthe ype s' Hence mark the region about the line (2) which lying towards the origin, , Constraint (3) is ofthe type "2. Hence mark the region about the line (8) which i lying away from the origin. ‘Thus, the common region isthe segment AE online (2). A, Elieson ines (2) and (3) % até, y= 200 ss From(), x+200= 500 « x= 300 = € (200, 200) (©) Value of Objective Funetion C = Sx + By at the comer points: ‘at A(Q, 500), Ca = 540) + 500) = 4000 ‘at E(300, 200), C: = 5(300) + 8(200) = 3100 (@) Optimum Solution: We have, Cr = 3100 at £(300, 200). Thus, the optimum solution is to mix 300 kgs of F, and 200 kgs of F; forthe mixture so that the cost is minimised 10% 3100. Problem 12: Solve graphically Minimise 2= 6a + hy Subject to Sey + 4x 2 60 Bat Te s 84 wt De 2 18 m2 0 (P.U. MBA - Dec. 11) Solution: 1. The Constraint Lines are: (5x4 4y= 60 fl) Putx:=0, «4p = 60, sx = 15 2 A(0, 15) isa point ont Puts = 0, «. 5x = 60, <+_B(A2, 0) is a second point on the line. ss Se ee ae 156 DECISION ScHNCES tar cnt ei 3 + Pe = 84 , +. (0,12) + €(28, 0) -@) Fig. 114 the type '2'. Hence mark the region lying away from origin. ‘the type ‘s': Hence mark the region lying towards from origin. ‘of the type '2', Hence mark the region away from origin Hence, the common region isthe feasible region GHFD, As G lines on lines (1) and (2), 5a + dig = 60 ~@) (3 gives) 15x + 12g = 180 Also, Bat Tx = 84 @ («5 gives) 15x + 35k = 420 and we have seen above 15q + 12e = 180 Fr Scanned with CamScannerLINEAR PROGRAMING peousmgs 9 ‘ «rom sea) wo Sx, = 60-33 ° 23 in H lies on (2) and = xi 2te = 18 0) («2 gives) 3. Value of Objective Function z = 6x: + 14% at comer points: ‘24 240) a) (240). 3864 no (39) zo dB) GG) Se atH (8, 5), my 145) = 218” Zr © 6{8) +140) = 108 Zp = 6(28) + 14(0) = 168 FAB, 0). Thus, the optimum solution is xi = 18, x; = 0, where Problem 13: A manufacturer of furniture makes two products chair and tables. The ‘manufacturing of these two products is dane on two machines A and B.A chair requires two hours on machine A and si hours on machine B.A table requires § hours on machine A and three hours on machi 20 and from table is € 0, Machine A is for 54 hours in a week. Solve the problem 1ot more than 9 tables ae tobe produced. (P.U. MBA - Dec. 12) i Let_ x Number of chats produced ina week yy ~ Number of tables produced in a week 1. Objective Function: P piece, Hence, total profits (20x + Maximise, pa con srENCES () Machine A: Each chair requires 2 hours on machine A vii Luncan programs each tables requires 5 hours on machine A. Hence, machine A is required for (2x + 5y) number of hhours. The aval Sys 50 Machine B: Similarly we get, +3y 5 54 Production of tables: This cannot exceed 9. Hence, yso 3, Thus, LPP model is Maximise P= 20x + Soy Subject t0, 2+ 5y < 50 6x +3y 5 54 ys9 xy20 2x-+ Sy = 50 + Points A(O, 10), B(25, 0) + 3y = 54 Points C(O, 18), D9, 0) -( ay of Ais 50 hours a week. Therefore, we must have, Scanned with CamScanner» LUNEAR PROGRAM DECISION SCIENCES are ofthe type = Hence, mark the “tus. the common region ODGFE is hy (b) Feasible Region: All the constraint about them, lying towards the origin. feasible region. Point tes on lines (2) and (3) Hence, solving et 25,9) Pn eon ws 1) an Hees sig te est we get GIS.) (a vant obec neon? = 20+ 50 ae pons Po = 20(0)+ 50(0) =0 Po = 2049) + 50{0) = 180 Pe = 20(5) + 5048) = 500 at F25, 9), Py = 20(25) + 50(9) = 500 at (0, 9), Pe = 20(0) + 5018) = 450 {(@) Optimum Solution: We see that Pra = $00 at Fand G. “Thus, the problem has muiple solutions which correspond to the coordinates of infac any point lying on the section PG of line (1) these equations simultaneously 3s simultaneous, P [LHS of object function line 20x + S0y = P ie. 2x + Sy = Zp and that of line (i same ie. they have same slope, Thus, the farthest iso-profit ine wi Now, as x and y represent produced chairs and tables, they must be integers and not fractions, We see that point G(S, 8) is the only point on Section FG, having intege iven at Gi. produce S chairs and 8 tables whid Problem 14: Solve the following LPP graphically (graphical method) Maximise, Z=120x+ 100 y Subject to the constrains: 10x + Sys 80 6x + y 5 665 fe By 228 5x+ 6y < 90 x20, y20 (S.P.P.U. MBA - May 3) eo —SS—S Lees DECSION ScuNCES Bsenses ves rcceanis Solution: 1. The constraint lines are: () 1ox+ Sy = 80 Putx=0 Sy =80 Puty=0 +. 10 80 +, B(B, 0) is second point on the line (i) 6+ 6y = 66 Putx=0 .y=M + ¢(0,1) Puty=0 .x=n + DAL0) ax + By = 24 Putxe 0 2 a3 Puty=0 x26 (69 5x4 6y= 90 Putx=0 0 y=a5 Puty20 x=18 + G(0, 15) = HB, 0) Fig. 1.16 FT Scanned with CamScannerLUNEAR PROGRANN AY, DECSION SCENES ign. come onthe fel fd is coordinates atone 2) (Oviding by we Ge xtye e5.gives) 5x + 5y = 55 5, 10x + 5y = 80 a) Subtracting, -Se=-25 «x85 From @), 615) + 6y = 66 635 iey=6 ‘a redundant constraint) le of objective fraction Z = 120x + 100y at comer points ALCO, 10) Ze = 120(0) + 100(11) = 1100 ALE(O 3), 2 = 120(0) + 100(3) = 300 ALF, O) 2, = 120(6) + 100(0) = 720 ALBE, 0), Za = 120(8) + 100(0) = 960 \ ALIS, 6, 2 = 120(5) + 100(6) = 1200 4. Optimum solution: We have, Znac# 1200 at IS, 6) Thus, the optimum solutions x= 5, y = 6 Minimise Z=25x+ 30y Subjectto 4x+ 3y260 x 4 3y236 xyz0 (ans. x= 12, y = 4,2 = 420) eee 160 EIN SINCE Luvern procan (it) Maximise Z = 10x + 154 Subjectto <3 mss 3x, + xq = 29 420 (Ans. Feasible Region is only point (3,5) Soln x: = 3, x; = 5,Z = 105) Maximise Z=3x +5) Subjectto 3x, + 4x22 Dyn m2-2 2 + 34812 ¥S4,%22 X20 (Ans. x: = 3/4, x= 7/2,Z = 1975) (wv) Minimise = 6x+2y Subjectto 3x+y=3 6x + 3y26 xy20 (Ans, Multiple solution Soln.: section (1,0) to (0,3), = 6) () Maximise © 2 = 5x + 2x2 Subjectto x2-y arys2 yea xy20 (Ans. infeasible solution) 3. A construction company wants to transfer the construction material from its warehouse to the project site, using two special types of trucks -A and B. Truck A can carry L tonne of cement and 4 tonnes of stel in one trip and truck B cartes 2 tonnes respectively. How many trips for each truck be planned daily, so as to minimise the {transportation cost. Formulate the LPP and solve it by using Graphical method, (Ans. Let x, x2: No. of trips of A and B respectively. LP model is Minimise Z = 500 x + 700 x Subject to x; + 24226; 4x4 + 3x 2 12; 4 + %2S20; xym20, Soln.: x: = 1.2 = 2 trips (approx), x2 = 2.4 = 2 trips (approx) Cost = € 2,400) 1a Scanned with CamScanner4 LINEAR PROGRAMA wo products ~ chats and ta, ogany timber ‘400 board feet of mal on ote a char requires S board feet ang fable uses 20 board feet ang ans to make ‘A furniture manufacturing compary Pl labour. It knows th 1, wl 4 : Aa mn ous rd cae pee to arin Ht ANY, man-hours and Fas 3p tar vvrees constant 5 tha sedabies he come make keen gearing model and. provide graphical solution. produced. LP model i, = No, of chairs and tables mowed ug bjt» ax, 8; + $9.00 falnsn = 220) in the profit maxi necks inthe production process? vis ¥4-Nos. of X; and % produced PP Maximise Z = 20%, + 2512 Subject to 10x, < 5000; 15t + 10 $000, 3% $12,000; x, 92 0. (9 333 of X, 400 of Xe, Profit = € 16660, (i) Since quantity of Zz remaining ie, $000 ~ (15(333) + 10(800) « 5 is very small we have 2670 units of Zs abo unutiized) ‘A company producing three brands of Shampoos has two plants located at two aces. Each plant has following production capacities per day: ants Brands (Botles per day) Fresh Blossom ‘Moon T 3000 1000 72000 1 1000 1000 6000 ‘A market survey indicates that during any paricdar month there will be minima demand of 24,000 bottles of Fresh, 16000 bots of Blossom and 48,000 bottles of Moon, The operative costs per day of running the plants I and it are 600 monetay Units and 400 monetary units respectively. The other fed overhead 1 are © 100 and € 150 per doy respectively. How many day company run each plant duting the month so that the preduction hil stil meeting the market demand, Use Graphical method of selon (ars st mM. of ys las ad ae nn a month especvey. LP Minimise Z = 600 x, + 400s: + 250. Subject to 34, © +2 ih 4 + 32 24) m%y 20. ecto an + 2a +H 2 Soln: Xi = 4 days z= 12 days, Cost = £7450) Tee ecstouscences ean rrccranans 7 LuNeas PROGRAMMING Solve the following LPP. by graphical method: Maximise Z = 5x + 7x Subjectto x + x, $4, 34; + 812524, 104+ 70535, 26 (Ans. infeasible solution) A toy company manufactures two types of dolls, a basic version — doll A and a deluxe version - doll 8. Each doll of type 8 takes twice as long to produce as one of type A and the company would have time to make a maximum of 2000 per day, The supply Of plastic is sufficient to produce 1500 dolls per day (both A and 8 combined). The deluxe version requires a fancy dress of which there are only 600 per day available. If the company makes a profit of € 3.00 and f 5.00 per doll respectively on doll Aand 8, then how many of each doll should be produced per day in order to maximise the total profit. Formulate and solve LPP. (Ans. Maximise P= 3x + Sy st: x + y s 1500, Prox = 500 when xi.e, A= 10.and y ie. B = 5) $600, 2y + x < 2000; xy 2 0, 3. Solve the following LPP using graphical method Maximise Z = 4X, Suibject to constraints: 2X; + 3X, 2 30, X; +X $14, % #2X) $18; Xy Xz 0 (Ans. Zpuc® 52 at Xs = 12, Xe = 2) (P.U, MBA - May 06) Solve by graphical method Maximise Z = 6x + 14e2 Subject t0 Sty + 4422 60, 3x; + 7K 2 84,1 + 20218) H, 1B O (Ans. Multiple Solutions Zn = 168) (P.U. MBA May 07) ee A firm makes two types of furniture - chairs and tables. The contribution to profit by each product as calculated table. Both the products time requited in hours by ‘each machine areas follows: ‘accounting department is € 20 per chair and @ 30 per 10 be processed on three machines My Mz and Ma, The product and total time available in hours per week on Machine [Chair Table | Available Time (in hrs) q 3 3 36 Ma 5 2 50 My 2 6 0 How should the manufacturer schedule the production in order to maximise the (Ans. 3 chairs, 9 tables, profit = 330) (P.U. MBA - May 08) Fr} Scanned with CamScannerLUNEAA PROGEANAN, beasston sciences od 12, Salve the following LPP using repbical meth we nabbas 6) Manimise 7 = 5q + Dey Set 1 40 + 05 16 38 672 ss preteen (PU. MBA - Dec. 05) (Ans. x, = 1.89 = 6, Zan ® + Gi) Maxirise 2 = Sx + Ta Subject tom #1 £4 # Bt $24 208, THD 5 35 m2 0. (P.U. MBA - May 11) nding machines to produce two machine thes, milling machines and grinding mach r parts. Table given below represents the machine times required for each par, :chining times available on different machines and profits on each machine part Machining time required for | Maximum time ‘Machines machine part in a available (minutes) 6 73000 10 2000 3 900 100 (P.U. MBA - May 05) ser to spread out atleast an 7200 kgs of a special his fields. There are Wo sources of obtaining these miures. Mstue A and miture B bath are available in bags weighing 100 kg each and they cost € 40 and € 24 respectively. Minture A + contains phosphate and nitrogen equivalent of 20 kg and 80 kg respectively while ‘moqure 8 contains thse ingredients equivalent of 50 kgs each type. Determine how tmany bags of each type should the farmer buy in order to obtain the required a minimum cost (Ans. A =O, B = 144, Coy = € 3456) (P.U. MBA - May 23) 15, Maximise, Z= 50a + 30x Subjectte: 2 +m 2 18 nme 2 Bay +2 < 34 a2 0 Solve by graphical method (Ans. Zoos = 560 at xs = 10, x= 2) * sprue’ SSS ieee ECs ScENCES ‘mansponramion Moose Chapter &... Transportation Models Contents ... 2.1. Introd ear Programming Formulation of a Transportation Problom Terminology of Transportation Pe Solving @ Transporttion Problem Solution hod For Finding the Optimum Solution 2.7. Special Cases in Transportation 2.8 Note Solved Problems Exorcite RS Its a typical Operations Research technique intended to establish a least cost route for the transportation of goods from the sources (say @ company’s production plants) to the various destinations (say i's warehouses) A transportation problem is a special type of Linear Programming Problem which can be solved by th ‘method of solving these problems. However practically an altemate method called as transportation method is used to solve them, Conser a problem of transporting goods destinations (warehouses) W, Wo, Wp The capacities respectively while the expected demands at the respective warehouses are di indicates the cos of transporting one unit fom the source Fo destination W,. Thus Canis Per unit transportation cost from the second factory Fz to the third warehouse Ws, It is assumed that the total supply available atthe sources will exactly satisty the total demand required at all the destinations ie. + S2 +. Sy = dh + ds + # de 2a Scanned with CamScanner{RINSPORTATION Mopeis DeCSION sciences ‘of a transportation table, Soe oman cn et dein te om nr deaghaiew a oa is a tabular arangement of the data about the demands (4) 2 the dest he peru transprtaion o SHPING ost (C) avalablties atthe supply points (5) and between each supply point and the destination as shown below. Sources — Supply capacity wi | ow |. | Ww Ft ce Gu| 5 fr al» | | & Fe Cul Go] | Go| Se Demand | ch] |__| __ds|_ N= Total Supply/Demand ‘The problem is to find the number of units to be transported from each of the sources to each of the destinations so tha the total os of transportation 1, Objective Function: Now, let X; represents the numberof units tobe transported from source (=, 2.) ‘to the jm destination G = 1, 2, . n). Thus, Xap represents the number of units to be transported from the source Fo warehouse Wi. The cost of wansporting these units will be imised. (CaXn) a5 Cais the corresponding per unit transportation cost. 2 Total Cost Z = Guan # CuXnt m+ GaXint Cokin oe + Com Ken == Egy, ina 2. The objective function is to Minimise Za BGK it jet SS cision scices _eansponranion moves 2. Constraints: (0) Total supply from the frst source to all the warehouses must be equal to S; ie Xr tXnt ot Xin = Sy ie Ixy =, ja for each source. Hence in general we have, EX; = § foreachi=1,2,...m = (i Simiaty, forthe fist warehouse Yan tka ta tea = dh ie. EX = dy it Hence in generat © Xj = djforeachj=1,2,..0 ‘Similarly, the non-negativity condition is 52 Ofor each i= 2,2,..,mandj = 1,2 ‘Thus, the L P. model forthe Transportat Minimise Zed BGK inh jot «(Cost minimisation) EX, = 5 foriea ia Subject to (Supply Constraints) EX +4 fost an 420 (. Non-negtity Condition) This problem can be solved by using the simplex method of solution. But for a transportation problem we us the transportation method which s moce convenient to use. Note thatthe Total Supply (S; + S: +... + Sql = Total Demand (6 + 6 +... + 69 which is called the rim condition. (). Balanced Transportation Probl the total capacity ofthe supply points or sources is equal tothe total demand atthe destinations. (Gl) Unbalanced Transportation Problem: Here the total supply is not equal to the total demand. «(Demand Constraints) Scanned with CamScannerTRANSPORTATION Mopets th (ii) Transportation Table: It is used to represen {from each source to each (emanation and per nt wonspotaion co destination Worohouse Pa Dakt Pune] Bancelore | Supply Mumbat : : ‘| 1000 um 7 & | @®D cr 10 4 cron | oy | - | @® +200 7 € rouse | ay") - "| aD - Demand 700 700 1000 | Tort = 3000 eg, The above table has 3 rows for three sources, 3 columns corresponding to 3 warehouses along with one row and column indicating the demands and supply. the figures in circles indicate the number of unis to be transported. Eg 300 unis to be transported from ‘Mumbai to Delhi where unit cst of transportation is € 10. We have to identity these figures to obtain the optimum solution. (Qv) Dummy Source or Destination: This is represented by adding an extra row or column to the transportation table with '0' per unit cost for each of its cells. tis used to balance an unbalanced problem by considering appropriate supply oF demand forit. (W) Initial Feasible Solution: It is a solution that satises the supply and demand conditions and yet it may or may not be the optimum one. There are three methods ‘of obtaining this soltion viz. North West Commer Method (NWCM), Matrix Minime or Least Cost Method (LCM) and Voge's Approximation Method (VAM (9 psi Staion fail solution which also gives a transportation plan wit minimum tea cost. Ths canbe obtined by using the Stepping Stone Method oF Mead Douton Methed (MODI Method, once we get the nal Feasble DeASION scieNceS TRANSPORTATION MODELS Ba Ganeral Procedure tor Solving a Transportation Step 1: Define the objective function to be minimised with the constraints imposed on Step 2: Step 5: ‘Step 6: step 7 (A) North - West Corner Method (NWCM): the problem. Set up the transportation table with m rows representing the sources and Columns representing the destinations along with their supply and demand figures as well as the unit cost figures in the cells. Add dummy row or column, if necessary, to balance the problem. Develop the Initial Feasible Solution to the problem using either NWCM, LCM or VAM. (of the intial solution. The solution is fea S at (men-1) number of cells at independent posit independent occupied ‘Test the solution, if found feasible, for optimality by finding out the opportunity costs forall the unoccupied or empty cells. Non-Negative (ie zero or positive) ‘opportunity cost forall empty cells indicates that the solution is optimal, If the solution is not optimal we improve it by using the Stepping Stone Method or the MODI method, Repeat steps 4, S and 6 until an optimal slutio Here we go on making maximum possible allocations to the North - West commer cells as follows: Step 1: Step 2: Step 3: Step 4: Select the North - West comer cell in the transpertation table and allocate as ‘many units as possible to it after checking the supply (in row) and the demand (in column) position for Reduce the supply and demand figures for the corresponding row and column accordingly. Cover the row or column where the supply or demand gets fully exhausted (Le. becomes 0) to get a reduced transportation table. Go to step 1 and repeat the procedure until total supply is fully allocated to the total demand. Note: Though it is easy, this method requires many further improvements to obtain the ‘optimum solution. 2 2 Scanned with CamScannerHANSFORTATIN Mode Beasousemiers ss dead ae wth sippy. demand and ung Illustration 1: Consider the following transportation oe cost figures as shown: - Dy | _sueely D 4 20 A 3 z - 30 5: 2 f 4 Ss 3 2 se : a 2 - 2 Demand | wo | 2 | | % Solution: This is a balanced table with total demand = total supply = 225. 9) Choose the North - West comer cell i. SDs, The row supply fs 20 and column demand is 40 for this cel, Hence we can assign at most 20 units to (id) After assigning these 20 units, the supply of $; becomes (20 - 20 = 0) and demand for D; reduces to 40-20 =20, Thus, cover row S; now as its supply has been full used, to get a reduced table. a TT i i a4] 7 3 @ 7 ys Demand | #620 [30 | as 2. Inthis table go again to the north west corner cll i.e. $,0, and allocate 20 units to it 2s it has 20 units of demand and 50 units of supply available. This reduces supply of row S; to 50-20 = 30 and demand of D; 020-20 = 0. & To To Ts S {ee 4 4 | 50,30 Demand | wa60] | es Cover D, as its demand has been uly satisied to get a new ted luced table, TOT Ee pecaton scence "TeansroRtation MODELS 3. In this table make allocation to the northwest corner celle. D3 of 30 units, This reduces both the demand of D, and supply of S to 0. Hence, cover both S) and Da. a Tb | Suey gots 3 OL 3 7 Teer [as Demand wots 4, Now, only one column D3 with $303 and S,D, remains. Make allocations directly to "them of the remaining 30 and 25 units of ‘demand for 55 units of Thus, here we are getting a plan to Dy, 30 from S; to D; ete (m+n-1 = 4+3-1 = 6), the solution isnot feasible. Least Cost Method (LCM) or Matrix Minima Method: Here we go on making allocations to the minimum cost cellin the table. i: (). Selecta cel the table. ith minimum unit cost i, there is a tie, then select that cell among them where more number of units can be allocated (after considering their row supply and column demands) If there ie again, then select a cell randomly from them. Step 2: Allocate maximum possible number of units to it. Reduce the corresponding supply and demand figures accordingly to get a reduced transportation table as in case of NWCM., Step 3:_Repeat steps 1 and 2 until entire supply is exhausted to fulfl the entire demand. lustration 2: For the same problem asin lustration 2. 1. @ Choose the cell with minimum unit cost. There are two such cells $:D, and 5,0, both having unit cost 1 i.e. there is. te, (id) Hence select that cel where more units can be allocated, After considering the row supply and column demand for these cells we see that we can allocate 20 Units to $,Ds and 50 units to $,0s, Hence select $0, 2. Allocate 50 units to it and then reduce its row supply and column demand by 50. ‘Thus, the supply ofS; gets fully exhausted and demand for D3 becomes 5. Hence, cover S;t0 get the reduced table, Ey Scanned with CamScannerTRANSSORTATION MoDRas sees aneronanon nous DECISION SCIENCES. pecs m aa a a Tet cot = 350) +500 +390) +1. «t= 35 = Sr ‘Ae, since there are m+ n-t » 4+ 31 = 6 number of Independent ocuped ces : the solution's feos at (©) Vogel's Approximation Method (VAM): ot aes Here we goon making allocations to the minimum cos cl ofa rw or column for which % ate theperaly for not making an aloction the opportunity co) gh Tend | jifference between the two smallest unit cost h row and colurn, column with highest penalty and choose the cell with the call with minimum cost is oF the reduced table we repeat the procedue C. » Gs Hence we ate aocaton oof units after coring m SPY an remaining column demand. We then reduce the row supply and clerin Ceriand ty 5 and get the new reduced tale by covering coulmn D, for and becomes 0 : a a rl ais the selected cell and reduce its row supply Scanned with CamScanner ‘and colirm demand accordingly, Obtah the feuced table then as dee ‘ @ previously. * Lo = = ‘Step 4: Recompute the penalties for the reduced table. [If a row (column) has been a, a © 7] ‘covered the row (column) penalties remain unchanged]. Repeat the above Benang | [sak rocedure unl the ene era and supply ets ctheusted we make allocation tothe minimum cost cele $0, of 5 9 Romy ate ns en Pe ees ons eae aed 1. Considering the same problem asin tlustation 1 and 2we compute the penahies for q ie rae ke allocation of 30 units to each row and column as follows: The penalty is obtained by taking the difference et Seen Se Se IR a .smalestunt cas gues pearing in each ov ad eon aie Ds D: Ds | Supply | Penalty emmy - St a wv A 20 2-1-1 50. St - 7 50 2-121 7 8 Ss 3 i x 30. 3-251 a oT [tT as Se y * 7s 6-452 fe 0.10 96:15 Demand 40 30 Es : 5 Sz we male slocatont ofthe remahing 10 and 15 unis nil PR [ae Loca feasible solution as follows * EE 2fateo D, BT ‘Supa Tere 2.) eniy the row/olu wih highest pena. These are row S and column 72S both wh penalty 2ie there's ate = Gi) From these two, select the one containing minimum cost cell: S, has minimum @ 5, 2) a cost cell Sz D; with cost 4 and D; has the minimum cost cell S, D3. with cost 2. Hence we select D>. 1 Gi) tn Dythe cell with minimum cost i.e. S, Ds is selected for making allocation. 28 28 _ OU KOLA EES Demand [40 | 30 | 5"TRANSPORTATION Moe prosowscmnes aces sets THansroRtanoN MODES ‘ow supply and column demand) ang er consideing row uP and err ean 6, We then make allocations to the reining row Sy directly to get the over $351 SUPPIY 9 » ee iy 10 9 feasible eth ede we ofS 55 Tay > BT eT 8 mea = =. =p 60} sa rs 3 77 2L =| a e 7 2B 2 40 20.10 = 7 2 [2 jes As a row has been covered, the 4. For the reduced table we recompute the pet tow penalties will remain unchanged while only the column penalties will change, as shown below Supply] Penaty 5 = [4 Ea inf 400 300 400 200 160 50 80 70. 140 340 jon: This a balanced table with total demand = total supply = 340, (@) North - West Corner Mi Now, the row Si: has highest panaty, Hence we select t and make maximum 2 allocations of 25 units tothe mi it ie-t0., Dj, We obtain thena Steps Select the north west comer cll ie Wn theta 7 supply is 70 and bseririeceyemtenaney column demand is 0, Hence, we can allot maximum of 50 units it Simeon Brae rovis and columns have penaky _-—-St®P2:- Allo SO units to ,Ws and hence reduce its row supply and column demand by of 1 We select row ing minimum cost eel $03 with cost 50. This makes the row supply 20 while the column demand gets exhausted as Then we allocate 50 units tot which reduces supply of $100. aoe Sa ee Step 3: As column demand of Wy gets exhausted fully, cover it to get the reduced en table as below: \ a oa We we [ws Tw s tL} 2], -= ft meh 3p BI Be F700] —s0o) —z00) —ee} Domand | 4615 | a610| 5 ny “155, 105] 00] 206] a0 Penal w]e trys 2 jo [_» | 7 | wo Scanned with CamScanner"RANSFORTATIN Mogg a table: 2. Now repeat cc ie “wot sal na hsted Hc, cove THE TOW P10 re s 2 ‘sat at y6 ato wo oy as aa} | 3. Repeat the procedure similarly for maki (of 30 units to P; Ws to get the following table. } We 5 oS 50} Ge El poo ?, 7, 2100 ° a] a | T Since now only one row is eft, make the remaining allocations to its cells directly to ‘ {get the initial solutior ees % oa 70 Pa QO] DOT 00] 0 Fe [alla a eo o [ata l[n| wl Here, Total Cost = 50(190) + 20(300) + 60(300) + 30/400) + 40(400) + 140(200) = 89,500. Note: All the operations can be carried out carefully on a single table to obtain tt solution. () Least Cost Method: Here we go on making allocations tothe least cost 1. Step’: Identify the least cost cell in the 3W both with cost he table, fare two such cells PW! thc le Hence ve sl hat ce i more numberof aos Considering the row supply 2 Column demand for these cls we see that we can alot 70 units 10 ANE ad 80 uso P.Ws Hence we see ie TT oS DeCSION ScHNCES ‘ansponraTion ODEs 2, Repeat the procedure for the o Step 2: We allot 80 units to P,Wy and then reduce its row supply and column demand by 80. This makes the row supply (180 ~ 80 = 100) while the entire column demand gets exhausted. Thus, we cover column W; to get the reduced table as follows: WwW, We W, Ww. s o ‘sot | 39 5 4 3 ae) — te} ae} — B00) Bl ep] a wi ops seo | 7 | 10 tT to the least cost cell teduced table and make second allocation (of 70 units) we Tw Tw s z San} — mn Pe] 78 [soo] 400] 60] oo Py 0) Spree] — Foo} —ze 0 > | ® 7o_| wero Inthis reduced table te least cost cells PW, Therefore make the thie allocation to it a a 7 Pi 7am] [S09] 0 F 0 co Py Gp fanaa o [= [ & | 7 | wo T fourth allocation tothe least cost cell Ps 1. Now in the new reduced table make th 25 70 units canbe alloted toi (whch s more than 30 units that can be allocated to PAW or PW Wy We W We s P| 7 0] apr) | 0|_saa Ps x x x [0.0.30 o | = | & | #e [te + a Scanned with CamScannerHANSPORATIN Gp, ein scences cals ofthe remaiting COMMA Wy sons diet 10 the Now, making further alcation direct ws Baw we, eo 3 | or we 100, 10 30 400) + Here total cost = 70 (100) + 20,700) + 70(400) + se aoe obtained by using one table only the operations are cai ‘Approximation Method: i oneduna Heer eee nang alocavons to the minimum cost cel in the row or column i highest penalty 1. Step: We co Fe] ae ||| ao _| #00-9008900 wag] Tea [—700-[ 250 | 80 | a0-100=100 2 |e |» | | 0 Penalty io 0-210] so cz0| saat [0-009 fh Step 2:_ Since the maximum penaly (210) occur for Ws, we select column Ws, Tht PW (with cost = 190) (1* allocation) to the cell PW, and then reduce its row supply and colume ‘demand by 50. Thus the row supply reduces to (70 ~ 50 = 20) and bt column demand g lly exhar as 50 ~ 50 = 0. Hence, we cover t, column W) to get a reduced table as flows, “Tw Tw 1s Fi a Too 20 Ps i] a a Py To) a a D w_ [7% [i DECISION SCIENCES "TRANSPORTATION MODELS 2, For the reduced table we repeat the procedure: () We recompute the penalties. As @ column is covered, only the row penalties will change as follows: Ke [ow 7 wT surely a wo] [600] gy | e00-360 100 foo [400 [220 [a6 100 [200-100 =100 pho [7 | 140 |e | 100 ry (i) Maximum penalty (200) oceus for row Py as well as column W, Hence there isa tie, Hence, we go for the one which contains @ minimum unit cost cll, However, here also there is a tie as row P, contains cll P,Ws and column Ws contains cell PW both of which have the same minimum cost of 100. Now, we selec the cel, here maximum allocation is possible. We see that we can allocate 20 urits to PAW, and 80 uni column Ws for the second allocation and allocate 80 units to it. This reduces its row supply to 100 {and exhausts its column demand, Thus, we cover the column Ws to get a new ‘reduced table as shown above. 10 PyW;. Hence, we select c We repeat the procedure once again for the new reduced table. We recompute the Penalties and make allocation (3%) of 20 units to cell P,W, with (cost=100) in row Py, ‘with (highest penalty of 400) as < qi Ye [ws TT suppty | Penaiy So rn Sc ve mre 70 | [300] «>| 600] oo | amo wo] GD] | OT we s00 | 200 7 | ho ° | 00 ‘This exhausts the row supply of P, and we get anew reduced table as above, 2as Scanned with CamScannerTRANSPORTATION Wop, DECISION SCENCES penalties AS 2 TOW is covereg the 4. Repeat the procedure once again by recomPung only the colar peates wil change. , - a m_] supe [Penaty 7 | a mm ee = 300, 400 600} 90 200 Py ‘ a ae ole 28 Denana] vo | whale erally 7] t “Select here column We with highest penalty 400 and then make allocation (4°) toc P,We (vith lowest cost 200) of 100 units which exhausts the row supply of Ps ’As only one row is left now, we make the further allocations ofthe remaining units a its cells and thus get the intial solution 26 ae eee et pa ea aes Here, Total Cost = $0(190) + 20(100) + 70/400) + 20(600) + 80(100) + 100(200) = 79,500 ‘above tables, the numbers 1, 2 etc with the arrows near the penalty figure: the number ofthe allocation (1, 2" etc) that is made in that row or colunm number of the allocation after which, column gets exhausted or covered. (i) All the operations can be carried out ona single table carefully to get the solution @ save time as done ahead. ‘The NWCM, LCM and VAM ma 216 |Methed for Finding the Optimam Sglution solution is obtained we by fir dependent positions) led as a case of degen further using ‘optimum solution, use, only this method has been described here, a Decision scrnces ansronraTion MODELS (A) Modified Distribution Method (MODI) or UV method: ‘Step 1: Determine the Initial Basic Feasible Solution and verity that tis feasible Le it has (m#n-1) independent occupied cell Step 2: Determine the row numbers (denoted a5 u, ie. uy, up ete) and the column numbers (denoted as vi. vy vs ete) by using the formula: C, = u, + v for each ‘occupied cell (C; is the corresponding per unit transportation cost). Choose here one of the values of u, or v, as 0 arbitrarily (prefer that row or column which contains maximum number of occupied cell), and then use the formula for finding out the remaining value one by one. ‘Step 3: Now consider the unoccupied cells and find the opportunity cost {improvement index) for each of them by using the formula, Opportunity cost ay = Cy= (ui + y) for these cells Step 4: Check the sign of all the opportunity costs. If all of them are non-negative (ie zer0 oF positive) then it implies that the solution obtained is optimum If atleast ‘one of the opportunity costs is negative, then it implies thatthe solution isnot ‘optimum and there is @ scope for improvement, Then go to step 5. Step 5: Select the unoccupied cell wth highest negative opportunity cost if there are ‘more than one such cells then select the cell where mare units can be shifted, as described in further steps) Step 6: For this cell trace a closed path using most direct route through at least 3 ‘occupied cells and then back to the unoccupied cell Use only vertical and horizontal ines and take turns only atthe occupied cell. Step 7: Assign (+ve) and (-ve) signs alternately to each of the comer cells along the closed path, starting with a (+ve) sign for the selected unoccupied cel Step 8: Find the maximum number of units to be shifted to this unoccupied cell (These are equal to the least of the n Units in the cells with ve) signs). Add Units to the cells with (+ve) sign and subtract it from the cells along the closed path. ‘Step 9: Go to Step 1 and repeat the procedure until an optimum solution is obtained. ‘Note: (The inital feasible solution must have atleast one allocation in each of its rows and columns. (i) The allocations must be made in independent postions (ie. the positions of cells where a closed path can not be traced as explained above) and must be (m + m1) in number. ‘The opportunity cost (4 for an unoccupied ct if one unit is shifted to this cells. Hence a (-ve) opportunity cost indicates saving in the cost and hence we select the unoccupied cell with highest (-ve) opportunity cost. the total cost 27 Scanned with CamScanner