Introduction to Operations Research (OR) and Linear Programming (LP) Session-1 By Prof. Prashant PremkumarManagement Science and OR ° Broadly, the tasks of the manager are: © Handling Administrative Task ¢ Problem Solving © Strategy and Innovation ¢ Networking and bonding © The aforementioned tasks require ¢ Formalization © Quantification ° Constraints handling * Decision making © OR is about formal techniques, quantification, decision making and constraints and limitation handling,OR and LP © ORisa field of study where we optimize performance under given constraints © OR also helps us develop models for decision making © Among the several topics in operations research, LP is the most important and the most popular one. © LP problems have applications in the various fields. © Some of which include manufacturing systems, public systems, business supply chain management and analytics, expert systems and decision support systemsDefinition of Linearity © Linearity or Linear functions are those whose graph is a straight line © y=f(x)=at+bx © In LP, we will be dealing with linearity, i.e., each variable will be having power one. © In LP we often use inequalities. © An inequality is a relation which makes a non-equal comparison between two numbers or other mathematical expressions. ° Example, <, >, #, S,2To begin with LP: Example-1 © Formulation-1 Product mix Problem: ° A shop can make two types of sweets (A and B). They use two resources flour and sugar. To make one packet of A, they need 3Kg, of flour and 3Kg of sugar. To make one packet of B, they need 3Kg of flour and 4Kg of sugar. They have 21Kg of flour and 28Kg of sugar. These sweets are sold at Rs. 1000 and 900 per packet respectively. Find the best product mix to maximize the revenue.© What is the objective? ¢ To maximize the revenue © What decisions will influence the objective? © Quantity of Sweet A to be produced © Quantity of Sweet B to be produced © What resources do we have to implement the above decisions? e Flour © Sugar Are the resources unlimited?Let X1 be the number of packets of sweet A made Decision Variabl Let X2 be the number of packets of sweet B made. —————* version Varlables Res s sources Flour Flour 3Kg 3Kg 3X,+3X, < 21 sassy BIEG) 4Kg Sugar ccna 3X,+4X, < 28 _—_ X,,X,20 —— Non-negative Revenue S restriction Maximize 1000X, +900X, SN Objective FunctionCharacteristics of LP © Proportionality: © If make 1 packet of sweet A then I get 1000 rupees. If make X1 packet of A then I will get 1000X1 rupees ° Additivity: © Additivity property says that profit obtained by selling X1 packet and X2 packet can be added. ° Certainty: © We need deterministic values to formulate LP. E.g., 21Kg flour, 28 Kg sugar and so on. © Divisibility: © Since we are using continuous variables, the LP model assumes that the decision variables can take on fractional valuesLP Problem formulation: Example-2 © A company wants to advertise their product in four different media- TV, newspaper, website and radio. The reach per advertisement on these four platforms are 8000, 5000, 3000, 2000. The cost per advertisement is 4 lakhs, 3lakhs, 2 lakhs and 1.5 Lakhs. The maximum number of advertisements the company wishes to have in each media is 3, 4, 5, 4. The budget of the company for the advertisement is 32 lakhs. How many advertisements does the company decide in each media to maximize the reach? ¢ X1 be the number of advertisement onTV © X2 be the number of advertisement on newspaper © X3 be the number of advertisement on web ¢ X4 be the number of advertisement on radioLP Problem formulation: Example-2 (Cont.) © Maximize 8000X1+5000X2+3000X3+2000X4 © 4X143X2+2X3+1.5X4 $32 ° X13 © X2<4 © X35 ° X44 © X1, X2, X3, X40LP Problem formulation: Example-3 © The daily requirements of nurses in a private nursing home is given by the following table: ii 8AM to 12 noon 12 12 noon to 4PM 15 4PM to 8PM 10 8PM to 12 midnight 8 12 midnight to 4. AM 6 4AMto 8AM 10 © The nurses start work at the beginning of the shift (e.g. 8AM, 12 noon etc.) and work for 8 continuous hours. What is the minimum number of nurses required to meet the daily demand?Let X1 to X6 be the number of muses who start Gein) work at 8AM, 12 noon , 4 PM, 8PM, 12 Mid-night 8AM to 12 noon 2 and 4AM respectively. 12 noon to 4PM 15 subject tor 4PM to 8PM 10 8PMto 12 midnight 8 X,+X, 215 12 midnight to4AM 6 X,+X,210 4AMto 8AM 10 X,+X,28 X,+X,26 X,+X, 210 X,+X,212 X,,X,,X,,X,,X,,X,20 Objective: Minimize : X,+X,4+X,+X,+X,4+X,LP Problem formulation: Example-4 © The demand for the products for two weeks is 800 and 1000. In a week, company can produce up to 700 units at regular time at 100 rupees/ product. It can employ over time and produce up to an extra 300 units in a week at rupees 120/product. The cost of carrying the product from one week to the next is rupees 15/product/ week. How that company should produce in order to meet the demand at minimum cost. ¢ Number of products made using regular time in week one - X1 ¢ Number of products made using regular time in week two - X2 ¢ Number of products made using overtime in week one -Y1 ¢ Number of products made using overtime in week two -Y2 © Number of products carried from week1 to week2 - Z1Subject to: X, <700 X, <700 Y, <300 ¥, <300 X, +¥, =800+Z, Or X,+¥,—Z, =800 X,+Y¥,+Z, =1000 X,.X3.¥,.¥%.Z, 20 Objective: Minimize :100X, +100X, +120Y, +120Y, +15Z,Subject to: X, <700 X, <700 Y, <300 ¥, <300 X,+¥, >800 Z, =X,+Y,—800 X,+Y,+Z, =1000 X,+Y,+X,+¥,—800=1000 X,+¥,+X,+Y, =1800 X,.X¥.¥, 20 Objective: Minimize :100X, +100X, +120Y, +120Y, +15Z, Minimize :100X, +100X,, +120Y, +120Y, +15(X,+¥, —800)LP Problem formulation: Example-5 e Three friends (A, B and C) start from P towards Q which is SKM away. They have one cycle and only one person rides the cycle at a time. A, b and C walk at the speed of 4, 5, and 6 KM/hour and can ride the cycle at 7, 8 and 10 KM/hour. How do they travel such that all three reach Q at the earliest time? © Let X1 be the distance Cycled by A in KM © Let X2 be the distance Cycled by B in KM © Let X3 be the distance Cycled by C in KMLP Problem formulation: Example-5 (Cont. ) e That means, © (5-X1) be the distance walked by A in KM © (5-X2) be the distance walked by B in KM © (5-X3) be the distance walked by C in KM e Further, x1 G—-xX) . _AfA Aw © Time taken by A = 7 4 5— © Time taken by B= +o e Time taken by C= a + Gx*3) ° 10 6 e X1+X2+X3=5LP Problem formulation: Example-5 (Cont. ) ° All person reach to the destination when the last person reach. © We do not know who is the last person. He can be either A or B or C. Suppose time taken by the last person to reach destination is u. © Thus, u>X1 + G-xbd 7 4 Ga X2) e >— Oru2 5 X3. (5—X3) > —+——— ° Oru2 76 6Do it on your own © The Mighty Silver Ball Company manufactures three kinds of pinball machines, each requiring a different manufacturing technique. The Super Deluxe Machine requires 17 hours of labor, 8 hours of testing, and yields a profit of $300. The Silver Ball Special requires 10 hours of labor, 4 hours of testing, and yields a profit of $200. The Bumper King requires 2 hours of labor, 2 hours of testing, and yields a profit of $100. There are 1000 hours of labor and 500 hours of testing available. In addition, a marketing forecast has shown that the demand for the Super Deluxe is no more than 50 machines, demand for the Silver Ball Special no more than 80, and demand for Bumper King no more than 150. The manufacturer wants to determine the optimal production quantities that will maximize his total profit.Maximize Z = 300X, +200X, + 100X, Subject to 17X, + 10X, + 2X, < 1000 8X, + 4X, + 2X, < 500 X, $50, X, $ 80, X, $150 X,>0,X,>0,X,>0Introduction to Operations Research (OR) and Linear Programming (LP) Session-2 By Prof. Prashant PremkumarLP Problem formulation: Example-4 © A company wants to advertise their product in four different media- TV, newspaper, website and radio. The reach per advertisement on these four platforms are 8000, 5000, 3000, 2000. The cost per advertisement is 4 lakhs, 3lakhs, 2 lakhs and 1.5 Lakhs. The maximum number of advertisements the company wishes to have in each media is 3, 4, 5, 4. The budget of the company for the advertisement is 32 lakhs. How many advertisements does the company decide in each media to maximize the reach? ¢ X1 be the number of advertisement onTV © X2 be the number of advertisement on newspaper © X3 be the number of advertisement on web ¢ X4 be the number of advertisement on radioLP Problem formulation: Example-4 (Cont.) © Subject To: e X13 e X24 e X35 ° X44 © 4X14+3X2+2X341.5X4 $32 e X1, X2, X3, X420 © Objective: © Maximize 8000X1+5000X2+3000X3+2000X4LP Problem formulation: Example-5 ¢ A financial advisor who recently graduated from IRMA received a call from a client who wanted to invest a portion of a $ 150,000 inheritance The client wanted to realize an annual income, but also wanted to spend some of the money After discussing the matter, the client and the adviser agreed that a mutual fund, corporate bonds, and a money market account would make suitable investments.© The client was willing to leave allocation of the funds among these investment vehicles to the financial adviser, but with the following conditions: At least percent of the amount invested should be in the money market account A maximum of only 35% should be invested in corporate bonds The investment must produce at least § 12,000 annually (ROL) ‘The un-invested portion should be as large as possible © The annual returns would be 11 percent for the mutual fund, 8 percent for the bonds, and 7 percent for the money market ° Formulate an LP model that will achieve the client's requests. Ignore transaction costs, the adviser’s fee and so on.Solution to Example Problem No.5 e Let \,, \,, V, be the amounts to be invested in mutual fund, corporate bonds and money market ° Non-negativity condition: Xi, \2, \3 20 » Atleast 25 percent of the amount invested should be in the money market account X3 / (X1tX2+X3) 20.25 » A maximum of only 35% should be invested in corporate bonds X2 / (X1+X2+X3) $0.35 » The investment must produce at least $12,000 annually (ROI) 0.11X1 + 0.08X2 + 0.07X3 2 12000 Objective Function: Maximize Z = 150,000 — (X1 + X2 + X3)LP Problem formulation: Example-6 e Three friends (A, B and C) start from P towards Q which is SKM away. They have one cycle and only one person rides the cycle at a time. A, B and C walk at the speed of 4, 5, and 6 KM/hour and can ride the cycle at 7, 8 and 10 KM/hour. How do they travel such that all three reach Q at the earliest time? © Let X1 be the distance Cycled by A in KM © Let X2 be the distance Cycled by B in KM © Let X3 be the distance Cycled by C in KMLP Problem formulation: Example-6 (Cont. ) e That means, (5-X1) be the distance walked by A in KM © (5-X2) be the distance walked by B in KM ¢ (5-X3) be the distance walked by C in KM © X1+X2+X3=5 x 5—X1 © Time taken by A ae B= X 2) ® Time taken byB="g> tS , G=X3) © Time taken by c= 8 io 7 SLP Problem formulation: Example-6 (Cont. ) © [8, 6, 5] © [2,4,9] e [5, 3, 6] © The earliest time when all three are at the destination is influenced by the person who reaches last. eu2z8,u26,u25LP Problem formulation: Example-6 (Cont. ) eur 41, G-*D 7 4 X2 ~~ (5—X2) a a X3 (5-—X3) °uz 10" 6 © X14+X2+X3=5 © Objective: ¢ Minimize u e X1, X2,X3,u20LP Problem formulation: Example-7 - Investment planning ° A senior executive of a public sector company recently quit his job under the VRS with a hefty packet of Rs 1 Crore. © Messrs. Dhana-chor Chit Fund Company has offered the following investment scheme for the benefit of such retired people. © This scheme is available only for the next six months (Encashment is possible on 180" day)Problem No.7 - Investment planning e "Invest a certain sum (in lakhs of rupees) in any month, invest half of that amount in the next month. In the subsequent month, one would get twice the amount invested originally in the first month” Returns received at the end of any month can be used immediately for reinvesting either as a fresh investment or as a follow-up investment © Formulate an LP model to maximize the returnsSolution to Problem No.7 © Let X, = Fresh investment on the 0" 30", 60", 90%, 120% day respectively (t = 1,2,3,4,5) (in lakhs of Rs.) © Noting that, in any month, the cash outflow must not exceed the cash on hand, the following linear programme is formulatedSolution to Problem No.7 Day 0 (Today): =X, <100 Day 30: ~+X, <100-X, 2 Day 60: %2+x, <100-x,-*1-x,+2x, 2 2 <100++1- x, 2 2 Day 90: sy, <100+— XX 2 2 2 Day 120: Xa, <100+ 244 x +4a_x, 2 2, 2 2 Day 150: %* — <1900+ 414424 *a Xs, 2 2 2 2 2Solution to Problem No.7 © On 180" day, the cash on hand must be maximum Day 180: 100+ st 474 x,-7342x © Hence the objective lancom: is Maximize z=100+ 42 42a 4 Rey Ae X, 20 ¥t=123,45.Problem No. 8 - Stock cutting © Amill cuts 20-foot pieces of wood into several different lengths: 8-foot, 10- foot and 12-foot. The mill has a certain amount of 20-foot stock on hand and orders for the various sizes. The objective is to fill the orders with as little waste as possible © For example, if two 8-foot lengths are cut from a 20-foot piece, there will be a loss of 4 feet, the leftover amountProblem No. 8 - Stock cutting... Currently, the mill has 350 20-feet pieces of wood on hand and the following orders, which must be filled from stock on hand Size in feet Number ordered 8 276 10 100 12 250 Formulate an LP model that will enable the mill operator to satisfy the orders with minimum trim lossSolution - Problem No & Pattern No 1 2 3 4 Pattern 1 2 3 4 FOF ND Pattern 8+8 8 +12 10 + 10 8 +10 10 12 enone coro Waste 4 0 0 2Solution - Problem No 8 © x, — Number of 20-feet pieces of wood cut using pattern 1 © x,— Number of 20-feet pieces of wood cut using pattern 2 © x; — Number of 20-feet pieces of wood cut using pattern 3 © x,— Number of 20-feet pieces of wood cut using pattern 4Solution - Problem No 8 © Subject To: © 2x, +x, +x, 2276 © 2x, +x, 2100 © x, 2250 ex, tx, +x, +x, S350 © Objective: ¢ Minimize: +x, + Ox, + Ox, + 2x, Es ES Ea © X1, Xp, X3, x, 20 = Elo fe colo -lo 2 1 0 1 >a. 2 a 0 0 2Introduction to Operations Research (OR) and Linear Programming (LP) Session-3 By Prof. Prashant Premkumar[8, 6, 5] Find the minimum among the three u<8,us6,us5 Maximize uProblem No. 8 - Stock cutting © Amill cuts 20-feet pieces of wood into several different lengths: 8-feet, 10- feet and 12-feet. The mill has a certain amount of 20-feet stock on hand and orders for the various sizes. The objective is to fill the orders with as little waste as possible. © For example, if two 8-feet lengths are cut from a 20-feet piece, there will be a loss of 4 feet, the leftover amountProblem No. 8 - Stock cutting... Currently, the mill has 350 20-feet pieces of wood on hand and the following orders, which must be filled from stock on hand Size in feet Number ordered 8 276 10 100 12 250 Formulate an LP model that will enable the mill operator to satisfy the orders with minimum trim lossSolution - Problem No & Pattern No 1 2 3 4 Pattern 1 2 3 4 FOF ND Pattern 8+8 8 +12 10 + 10 8 +10 10 12 enone coro Waste 4 0 0 2Solution - Problem No 8 © x, — Number of 20-feet pieces of wood cut using pattern 1 © x,— Number of 20-feet pieces of wood cut using pattern 2 © x; — Number of 20-feet pieces of wood cut using pattern 3 © x,— Number of 20-feet pieces of wood cut using pattern 4Solution - Problem No 8 © Subject To: © 2x, +x, +x, 2276 © 2x, +x, 2100 © x, 2250 ex, tx, +x, +x, S350 © Objective: ¢ Minimize: +x, + Ox, + Ox, + 2x, Es ES Ea © X1, Xp, X3, x, 20 = Elo fe colo -lo 2 1 0 1 >a. 2 a 0 0 2Graphical Method Maximize Z = 40X, + 36X, Subject to 5X, + 3X, 245 X, <8 X, <= 10 X,,X, 20Simplex method AG, 10) BG, 10) *Feasible Region "Feasible Solution. *Infeasible solution "Optimal Solution C8, 5/3)40X,+36X,=720 40X,+36X,=540Simplex method AG, 10) B(8, 10) C8, 5/3)Infeasible and Unbounded and Problems ° An infeasible problem is a problem that has no solution while an unbounded problem is one where the constraints do not restrict the objective function and the optimal objective goes to infinityFrom General to Standard Form Slack variable: Added to a S constraint 2xi+ 3x2 S50 2x1+3x2+%3=50 Surplus variable: Added to a 2 constraint 5xit 7x2 2120 5xit+ 7x2 -x3=120 Dealing with Unrestricted in Sign variables x1—URS Let x2, x3 20; x1 = x2 —x3Introduction to Operations Research (OR) and Linear Programming (LP) Session-4 By Prof. Prashant PremkumarExample Susan Frank is the operations manager of Arc Manufacturing Inc. She has developed a linear programming model to help her determine the product mix on one of the three production lines for Arc Manufacturing Inc.: maximize = Z = 15x, + 20x, + 14x; subject to 5x, + 6X, + 4x3 < 210 pounds per day (Material A constraint) Ox, + 8x, + 5x3 < 200 pounds per day —_ (Material B constraint) 4x, + 2x + 5x3 <170 pounds per day (Material C constraint) XyXq,X320 — (nonnegativity constraint)Example - Answer Microsoft Excel 16.0 Answer Report Worksheet: [Solver Example.xisx]sheets Report Created: 17-11-2020 10:34:56 Objective Cell (Max) Cell_Name Original Value Final Value SD$6_Obj x2 S48. Variable Cells Cell_Name Original Value Final Value Integer $c$3_x1 o (0 Contin ‘SDS3_X2 5 5 Contin SES3_X3 32 32 Cor Constraints Cell_Name Cell Value Formula Status Slack $D$7 C1 x2 158 $D$7<=$F$7_Not Binding 52 ‘$DS8_C2x2 200 $0$8< ding ° $D$9_C3 x2 170 $D$9<=$F$9 Binding °Problem 1 © Marketing manager of Arc manufacturing disagrees with Susan to go for a single pricing policy © He believes differential pricing policy — different prices for different set of customers - would help reap higher profits © So, instead of using a single number, the manager wants Susan to use a price range for every product © What issues does Susan face and help her address themProblem 2 As per the optimal solution provided by Excel Solver, the optimal product mix consists of only Products 2 and 3. Product 1 does not form a part of the product mix © Susan deliberates on the following questions: 1. What should be the bare minimum profit coefficient of Product 1 for it to be a part of the optimal solution? >. How much would the company loose if Susan decides to manufacture Product 1?Problem 3 © Forecasting team tells Susan that they can expect a surge in the demand of Products 2 and 3 (approx 25% of current production) in the coming days © Susan decides to ramp up the production and would need extra units of raw materials. © Upon request by Susan, the supplier who supplies B & C agrees to supply the extra quantity at the following pricing schedule. Should Susan accept it or not? © A: $1.2 per pound B: $3 per pound and C: $0.25 per poundSensitivity Analysis Enables the decision maker to determine how a change in one of the values of a model will impact the optimal solution and the optimal value of the objective function while holding all other parameters constant. Provides the decision maker with greater insights about the sensitivity of the optimal solution to changes in various parameters of a problem. Change(s) in the value of objective function coefficient(s) Change(s) in the right-hand-side(RHS) value of constraint(s)Microsoft Excel 16.0 Sensitivity Report Worksheet: [Solver Example.xlsx]Sheet6 Report Created: 17-11-2020 10:34:57 Variable Cells Final Reduced Objective Allowable Allowable Cell Name Value Cost __Coeffi Increase Decrease $C$3_X1 0-106 is 10.6 1E+30 $D$3_X2 iS 0 20 24 10.6 SES3_X3 32 0 14 36 Ae ‘Constraints Final Shadow Constraint Allowable Allowable Cell Name Value Price _R.H. Side Increase _ Decrease $D$7 C1X2 158 0 210 1E+30 52 SD$8 C2 X2 200 24 200 70.90909091 30 $D$9 C3 X2 170 04 170 30 120Range of Optimality ° The range of objective function coefficient values for which the optimal values of the decision variables would not change A value of the objective function coefficient that falls within the range of optimality will not change the optimal mix of the variables, although the optimal value of the objective function will changeMicrosoft Excel 16.0 Sensitivity Report Worksheet: [Solver Example.xlsx]Sheet6 Report Created: 17-11-2020 10:34:57 Variable Cells Final Reduced Objective Allowable Allowable Cell Name Value Cost __Coeffi Increase Decrease $C$3_X1 0-106 is 10.6 1E+30 $D$3_X2 iS 0 20 24 10.6 SES3_X3 32 0 14 36 Ae ‘Constraints Final Shadow Constraint Allowable Allowable Cell Name Value Price _R.H. Side Increase _ Decrease $D$7 C1X2 158 0 210 1E+30 52 SD$8 C2 X2 200 24 200 70.90909091 30 $D$9 C3 X2 170 04 170 30 120Range of Feasibility © The range of values over which the right-hand-side (RHS) value can change without causing the shadow price to change © Within the range of feasibility, the same decision variables will remain optimal, although their values and the optimal value of the objective function will change © Analysis of RHS changes begins with determination of a constraint’s shadow price in the optimal solutionShadow Price © Within the range of feasibility, the amount of change in the optimal value of the objective function per unit change of the RHS value of a constraintSensitivity Analysis..... © Maximization Problems © Minimization Problems © 100% rule © The 100 Percent Rule for Simultaneous Changes in Objective Function Coefficients: If simultaneous changes are made in the coefficients of the objective function, calculate for each change the percentage of the allowable change (increase or decrease) for that coefficient to remain within its allowable range. If the sum of the percentage changes does not exceed 100 percent, the original optimal solution definitely will still be optimal. (If the sum does exceed 100 percent, then we cannot be sure.) © DualityMicrosoft Excel 16.0 Sensitivity Report Worksheet: [Solver Example.xlsx]Sheet6 Report Created: 17-11-2020 10:34:57 Variable Cells Final Reduced Objective Allowable Allowable Cell Name Value Cost __Coeffi Increase Decrease $C$3_X1 0-106 is 10.6 1E+30 $D$3_X2 iS 0 20 24 10.6 SES3_X3 32 0 14 36 Ae ‘Constraints Final Shadow Constraint Allowable Allowable Cell Name Value Price _R.H. Side Increase _ Decrease $D$7 C1X2 158 0 210 1E+30 52 SD$8 C2 X2 200 24 200 70.90909091 30 $D$9 C3 X2 170 04 170 30 120Example for demonstration Minimize Z = 40x, + 36x, Subject to 5x, + 3x, 245 = Ss x, <10 Mig Mi, Be TDChange in S constraint Optimal solution: x1 = 8, x2 =5/3 Z = 380Optimal solution: x1 =9, x2 =0,Z' = 360 Change = 360 — 380 = -20e While solving the problems... Step 1: Ifitis a S constraint: - Increasing the RHS will relax the constraint and hence it may improve the objective function value - Similarly, decreasing the RHS will tighten the constraint and hence may worsen the objective function value Step 2(a): If it is a maximization problem: - Improvement (Z')=Z+Ab x Shadow price of the constraint - Worsening (Z’)=Z-Ab x Shadow price of the constraint Step 2(b); If it is a minimization problem: - Improvement (Z')=Z-Ab x Shadow price of the constraint - Worsening (Z')=Z+Ab x Shadow price of the constraintChange in 2constraint Optimal solution: x1 = 8, x2 =5/3 Z = 380B(8, 10) XS 10 Optimal solution: (8, 20/3) ¥1=8, x2 = 20/3, * Z = 560 Change = 560 — 380 = +180While solving the problems... Step 1: Ifitis a 2 constraint: - Increasing the RHS will tighten the constraint and hence it may worsen the objective function value - Similarly, decreasing the RHS will relax the constraint and hence it may improve the objective function value Step 2(a): If it is a maximization problem: - Improvement (Z’)=Z+Ab x Shadow price of the constraint - Worsening (Z’)=Z-Ab x Shadow price of the constraint Step 2(b); If it is a minimization problem: - Improvement (Z')=Z-Ab x Shadow price of the constraint - Worsening (Z’)=Z+Ab x Shadow price of the constraintWhile solving the problems... Step 1: Check ifiit isa $ constraint or a > constraint Step 2(a): If it is a S constraint: - Increasing the RHS may improve the objective function ~ Decreasing the RHS may worsen the objective function Step 2(b): Ifit is a > constraint: - Increasing the RHS may worsen the objective function - Decreasing the RHS may improve the objective function Step 3: Check if the problem is a max/min problem Step 3(a) If it is a maximization problem: - Improvement (Z’)=Z+Ab x Shadow price of the constraint - Worsening (Z’)=Z-Ab x Shadow price of the constraint Step 3(b) If it is a minimization problem: - Improvement (Z’)=Z-Ab x Shadow price of the constraint - Worsening (Z’)=Z+Ab x Shadow price of the constraintSensitivity Analysis..... Maximize Z = 4x, +3x, +6x, subject to Oranges 3x, + 2x, + x, $920 pounds Grapes 2x, + 2x, +2x, $900 pounds Pineapple xX, +2x, +3x, $930 pounds Peeling / Cutting 1.2x, +1.4x, +1.5x, <1260 minutes Mixing / Packaging x, + 2x, + x, $600 minutes X,.X,,x, 20 12a Xi- California mix Xo- Florida mix X3— Hawaiian mixSensitivity Analysis..... Allowable Reduced Objective Allowable cell Name Coeffic Increase Decrease $884 Quantities California mix 2 ~ SCS4 Quantities Foridamx SSS $054 Quantities Hawaiian mix 240 . Final Shadow Constraint — Allowable Allowable cell Name Value Price RHLSide_Inerease Decrease ‘SES Oranges Usage TO DSS ERO 50 S83 Grapefruit Usage 300 8 900 5 260 SES9 Pineapple Usage 930 7 30 420 50 SESIO _Peeling/cutting Usage 620 260 «+30 648 SES] Mixing/packaging Usage 450 © 600 1E+30 150Sensitivity Analysis... The equipment used for peeling and cutting must be replaced. The new equipment will have a capacity of only 1200 minutes. What impact will this change have on the optimal values of the decision variables and on profit? What would the unit profit on the Florida mix have to be before it would become profitable to produce? Ifmanagement had a choice of obtaining more oranges or more pineapples, which one should be chosen? Why?6 Sensitivity Analysis... Management has just learned that an additional 50 pounds of pineapples are on. hand. What will be the change in the objective function? Management is considering making changes that will cause the profit on the Hawaiian mix to be $8 per unit. Would this affect the solution? Will it affect the optimal value of the objective function? Management is considering a change in equipment that would result in increasing the profit on the Hawaiian mix to $8 per unit but result in decreasing the profit on the California mix by $1. Would these changes be within the range of optimality? If so, how much would the optimal profit change?Sensitivity Analysis... Minimize Z =38x, +19x, + 60x, subject to Protein 2.5x, + 4x, +3x, 2794 pounds Fibe 3x, +2x, +4x, = 300 pounds Fat 2x, +x, + 2x, 2600 pounds X,.X,,X, 20 Xi- Bags of raw material 1 Xo- Bags of raw material 2 Xs— Bags of raw material 3Sensitivity Analysis... B. Sensitivity Report Adjustable Cells Final Reduced Objective Allowable Allowable cell Name Value Cost. «Coefficient Increase Decrease $8$4 Quantities Raw Material) 292 ° 38 0 26125 $C$4 Quantities Raw Material 2 6 0 9 418 0 $054 Quantities Raw Material 3 0 2 6 TE#0 2 Constraints — "Final Shadow Constraint Allowable Allowable cell Name Value Price «= sRLH,Side Increase Decrease sts7"" Protein constraint Usage | 794 Oo. “4 SES8__ Fiber constraint Usage 908, ° 300 608 1E+30 re SES9. Fat constraint Usage: 600 29, 600 352 4015Sensitivity Analysis... Which constraints are binding on the solution? How do you know? What does the reduced cost of 22 indicate? How do you interpret the 608 in the allowable increase section of the fiber constraint? Determine the range of optimality for the objective function coefficients of x1 and x2 What does the shadow price of 19 reveal? What is the range of feasibility for the fat constraint’s RHS?4 Sensitivity Analysis... Would a decrease to 575 for the RHS of the fat constraint affect the optimal value of the objective function? If so, by how much? ‘The manager is considering reducing each RHS by 9 percent and may ask you to provide information on how that would impact the total cost. What would you need to do? Why?Sensitivity Analysis... Maximize Z =12x, + 10x, +15x, + 11x, subject to Material 5x, +3x, + 4x, + 2x, < 240 pounds Machine 6x, +8x, + 2x, +3x, < 240 hours Labor 2x, + 3x, + 3x, + 2x, $180 hours X,,X5,X, 20Sensitivity Analysis... B. Sensitivity Report ‘Adjustable Cells Final Reduced Objective Allowable Allowable Cell, Name Value Cost Coefficient Increase Decrease $854 Quantities Product | 0 +08 2 08 1E+30 $CS$4 Quantities Product I! 0 -86 10 86 1+30 $D$4 Quantities Product Ill 2 0 5 0.666667 7.666666667 SES4 Quantities Product IV n 0 1 nS 0.285714286 Constraints Final Shadow Constraint Allowable Allowable cell Name Value Price RH. Side Increase Decrease ‘SFST Material Usage wm. ° 240 1E+30 48 SFS8 Machine Usage 240 06 240 30 720 SFS9 Labor Usage 180 46 180 30 20Sensitivity Analysis... What is the optimal solution and its objective function value? The manager is concerned because none of Product 1 is called for in the optimal solution. At what profit per unit would Product 1 come into solution? Determine the range of optimality for Products 1 and 2 What is the marginal value of an hour of machine time? Over what range of machine time is this amount valid? Why the shadow price of material constraint is 0?6. Sensitivity Analysis... Manager can secure additional labor hours by the use of overtime. This involves paying a premium of $2 per hour. How much overtime can be effectively used, and what will the net total profit be if that amount of overtime is scheduled? If the manager can obtain extra machine hours and labor hours, which one should be chosen and how much should be obtained?