Introduction to Linear Programming

A B Raju

Introduction - LP
LP is a problem solving approach to solve managerial problems

Introduction - LP
LP is a problem solving approach to solve managerial problems Has extensive uses in day to day business functions

Introduction - LP
LP is a problem solving approach to solve managerial problems Has extensive uses in day to day business functions A manufacturer trying to satisfy all the demand with minimum inventory

Introduction - LP
LP is a problem solving approach to solve managerial problems Has extensive uses in day to day business functions A manufacturer trying to satisfy all the demand with minimum inventory Financial analyst trying to maximise the ROI with best picks of stocks and bonds for his portfolio

Introduction - LP
LP is a problem solving approach to solve managerial problems Has extensive uses in day to day business functions
A manufacturer trying to satisfy all the demand with minimum inventory Financial analyst trying to maximise the ROI with best picks of stocks and bonds for his portfolio A marketing manager trying to optimise his media mix to maximise ad effectiveness with the given budget

Introduction - LP
LP is a problem solving approach to solve managerial problems Has extensive uses in day to day business functions
A manufacturer trying to satisfy all the demand with minimum inventory Financial analyst trying to maximise the ROI with best picks of stocks and bonds for his portfolio A marketing manager trying to optimise his media mix to maximise ad effectiveness with the given budget A logistics manager trying to minimise his overall cost of transportation for shipping goods from various warehouses to a customer for a given order

Introduction - LP
So many companies use LP for various purposes

Introduction - LP
So many companies use LP for various purposes Hanshin Expressway corporation uses LP for traffic control

Introduction - LP
So many companies use LP for various purposes Hanshin Expressway corporation uses LP for traffic control Marathon Oil uses LP for gasoline blending and to evaluate economics of new terminal / pipeline etc.,

Introduction - LP
So many companies use LP for various purposes Hanshin Expressway corporation uses LP for traffic control Marathon Oil uses LP for gasoline blending and to evaluate economics of new terminal / pipeline etc., Eastman Kodak uses LP for manufacturing products in its various plants across the world

Introduction - LP
So many companies use LP for various purposes Hanshin Expressway corporation uses LP for traffic control Marathon Oil uses LP for gasoline blending and to evaluate economics of new terminal / pipeline etc., Eastman Kodak uses LP for manufacturing products in its various plants across the world GE Capital uses LP for optimal lease structuring

Introduction - LP
Timber Harvesting Model at Meadwestvaco Corp. Co. is a major producer of premium paper for books, periodicals, commercial printing and business forms Produces pulp and lumber Designs and manufactures packaging systems Is a world leader in coated board and shipping containers

Introduction - LP
Timber Harvesting Model at Meadwestvaco Corp. Co. is a major producer of premium paper for books, periodicals, commercial printing and business forms Produces pulp and lumber Designs and manufactures packaging systems Is a world leader in coated board and shipping containers Co. uses LP for Timber harvesting plans are developed over long time horizons

Introduction - LP
Timber Harvesting Model at Meadwestvaco Corp.
Co. is a major producer of premium paper for books, periodicals, commercial printing and business forms Produces pulp and lumber Designs and manufactures packaging systems Is a world leader in coated board and shipping containers Co. uses LP for Timber harvesting plans are developed over long time horizons
Models shall consider wood market, mill pulpwood requirements, harvesting capacities and general forest management principles Model arrives at a optimal harvesting and purchasing schedule Various schedules reflect other assumptions concerning forest growth, wood availability and general economic conditions Various factors like timber prices & supplies and land holdings & forest growth

Introduction - LP
So in all LP problems we have An objective of either maximising or minimising a particular quantity Constraints are another prominent feature of every LP problem.

Introduction - LP
Par Inc. wanted to produce standard and deluxe golf bags which have the following operations as summarised below

Production time (hours Department Standard bag Deluxe bag Cutting & Dyeing 7/10 1 Sewing 5/6 Finishing 1 2/3 Inspection & packing 1/10 1/4

Introduction - LP
Total production hours available for the next 3 months in different departments are
Cutting & dyeing Sewing : Finishing : : 630 hours 600 hours 708 hours and : 135 hours

Inpsection & pakg

Profit contributions per unit are

Standard bag Deluxe bag : $9 : $10

Introduction - LP
Problem is to be defined first properly for it to be solved. So the various steps are Describe objective Describe each constraint Define the decision variables

Introduction - LP
Problem is to be defined first properly for it to be solved. So the various steps are
Describe objective Describe each constraint Define the decision variables

For Par Inc, we can start with decision variables as

No. of Standard bags to be produced = S No. of Deluxe bags to be produced = D

Introduction - LP
Problem is to be defined first properly for it to be solved. So the various steps are
Describe objective Describe each constraint Define the decision variables

For Par Inc, we can start with decision variables as

No. of Standard bags to be produced = S No. of Deluxe bags to be produced = D Need to maximise the profit contribution, so the objective would be

Introduction - LP
Problem is to be defined first properly for it to be solved. So the various steps are
Describe objective Describe each constraint Define the decision variables

For Par Inc, we can start with decision variables as

No. of Standard bags to be produced = S No. of Deluxe bags to be produced = D Need to maximise the profit contribution, so the objective would be Maximise 10S + 9D

Introduction - LP
Constraint 1:

Introduction - LP
Constraint 1: Total hours for cutting & dyeing are fixed i.e. 7/10 S+1D 630 Constraint 2: total no. of hours for sewing 1/2S + 5/6 D 600 Constraint 3: total no. of hours for finishing 1S + 2/3 D 708 Constraint 4 1/10 S + D 135 Any other constraints???

Introduction - LP
Constraint 1: Total hours for cutting & dyeing are fixed i.e. 7/10 S+1D 630 Constraint 2: total no. of hours for sewing 1/2S + 5/6 D 600 Constraint 3: total no. of hours for finishing 1S + 2/3 D 708 Constraint 4 1/10 S + D 135 Any other constraints??? Non-negativity constraints S, D 0

Introduction - LP
Mathematical statement/model of the Par Inc. problem is Max 10S + 9D subjected to 7/10 S+1D 630 Cutting & dyeing 1/2S + 5/6 D 600 Sewing 1S + 2/3 D 708 Finishing 1/10 S + D 135 Inspection & packing S, D 0

Introduction - LP
Mathematical statement/model of the Par Inc. problem is Max 10S + 9D subjected to 7/10 S+1D 630 1S + 2/3 D 708 S, D 0 Why the above problem is called Linear? Cutting & dyeing Finishing

1/2S + 5/6 D 600 Sewing 1/10 S + D 135 Inspection & packing

Introduction - LP
Graphical solution procedure:

Introduction - LP
Graphical solution procedure: Graphical solution is possible only when two decision variables are involved

Introduction - LP
Graphical solution procedure: Graphical solution is possible only when two decision variables are involved Find out the solution points by on both axes keeping one variable 0 in the constraint equation assuming equality.

Introduction - LP
Graphical solution procedure: Graphical solution is possible only when two decision variables are involved Find out the solution points by on both axes keeping one variable 0 in the constraint equation assuming equality. A constraint line can be drawn using these two points

Introduction - LP
Graphical solution procedure:
Graphical solution is possible only when two decision variables are involved Find out the solution points by on both axes keeping one variable 0 in the constraint equation assuming equality. A constraint line can be drawn using these two points As the equation mentions , it is obvious that solution points falling below this line only will satisfy the equation.

Introduction - LP
Graphical solution procedure:
Graphical solution is possible only when two decision variables are involved Find out the solution points by on both axes keeping one variable 0 in the constraint equation assuming equality. A constraint line can be drawn using these two points As the equation mentions , it is obvious that solution points falling below this line only will satisfy the equation. So the all the four constraint lines can be constructed this way and they can drawn on one graph to see the solution area that satisfies all the constraints.

Introduction - LP
Graphical solution procedure:
Graphical solution is possible only when two decision variables are involved Find out the solution points by on both axes keeping one variable 0 in the constraint equation assuming equality. A constraint line can be drawn using these two points As the equation mentions , it is obvious that solution points falling below this line only will satisfy the equation. So the all the four constraint lines can be constructed this way and they can drawn on one graph to see the solution area that satisfies all the constraints. This solution area is called feasible region and all the points/solutions falling within this region are called feasible solutions.

Introduction - LP
Now we found the feasible solution area but what is the answer? The solution point which gives us the maximum profit contribution. On the boundary line, there are infinite number of points and hence trial and error method cannot be used to find out the solution. P = 10S + 9 D solving for D we will get D = 1/9 P -10/9S So from the regression equation like the above one, we can conclude -10/9 is the slope of the line and 1/9P is the D intercept

Introduction - LP
Using various values of P, we can draw the profit lines. As we move away from the origin, we get more profit and all the profit lines shall be parallel to each other as the slope is same Lets take examples of 1800, 3600 and 5400 profits and draw the lines. But we cannot go beyond feasible region as it is not acceptable because all constraints might not be fulfilled. As we draw the draw the parallel lines the optimum solution lies on the intersection of cutting&dyeing and finishing lines. By solving these two equations, we can find values of S and D

Introduction - LP
As observed, the optimal solution is the intersection of Cutting & Dyeing and Finishing constraint lines. So by solving these two constraint line equations 7/10 S+1D 630 and 1S + 2/3 D 708, we get D = 252 and S = 540 units. For this optimal solution the profit contribution can be obtained by substituting the values of S & D in the profit contribution equation 10S + 9D . So the profit contribution is $7668.

Introduction - LP
Slack Variables: Lets determine the production time required for producing the optimum solution quantities in various departments

Constraints Hours required Cutting & Dyeing 630

Hours available 630 600 708 135

Unused hours 0 120 0 18

Sewing 480 Finishing 708 Insp. & Pack. 117

These unused capacities are known as slack

Introduction - LP
Slack Variables: Often slack variables are added to linear programming problem to represent the slack or idle capacity. So using the slack variables the LP model can rewritten as Max 10S + 9D + 0S1 + 0S2 + 0S3 + 0S4 subjected to 7/10 S+1D + 1S1 = 630 Cutting & dyeing S + 5/6 D + 1S2 = 600 Sewing 1S + 2/3 D + 1S3 = 708Finishing 1/10 S + D + 1S4 = 135 Inspection & packing S, D, S1 , S2 , S3 , S4 0 LP when written as above with all constraints expressed as equalities, it is said to be written in Standard form

Introduction - LP
Extreme points and the optimal solution: Lets change the profit contribution of Standard bag to $5 from $10 without changing the constraints. Then only the profit function will change to 5S + 9D to be maximised and the optimal solution can be found out using the graphs again.

Introduction - LP
Extreme points and the optimal solution: Lets change the profit contribution of Standard bag to $5 from $10 without changing the constraints. Then only the profit function will change to 5S + 9D to be maximised and the optimal solution can be found out using the graphs again. Here we find when we draw the profit contribution lines, the optimal solution occurs at the intersection point of Cutting&Dyeing and Inspection & Packing constraint lines.

Introduction - LP
Extreme points and the optimal solution: Lets change the profit contribution of Standard bag to $5 from $10 without changing the constraints. Then only the profit function will change to 5S + 9D to be maximised and the optimal solution can be found out using the graphs again. Here we find when we draw the profit contribution lines, the optimal solution occurs at the intersection point of Cutting&Dyeing and Inspection & Packing constraint lines. By solving these two constraint equations, we get S=

Introduction - LP
Extreme points and the optimal solution: Lets change the profit contribution of Standard bag to $5 from $10 without changing the constraints. Then only the profit function will change to 5S + 9D to be maximised and the optimal solution can be found out using the graphs again. Here we find when we draw the profit contribution lines, the optimal solution occurs at the intersection point of Cutting&Dyeing and Inspection & Packing constraint lines. By solving these two constraint equations, we get S = 300 and D = 420 As profit contribution of S has reduced the quantity of production to be done also reduced.

Introduction - LP
So if we closely observe the optimal solutions lie on the intersection points or vertices.

Introduction - LP
So if we closely observe the optimal solutions lie on the intersection points or vertices. In LP terminology, the vertices are called extreme points

Introduction - LP
So if we closely observe the optimal solutions lie on the intersection points or vertices. In LP terminology, the vertices are called extreme points So while finding out an optimal solution, we need not evaluate all the feasible solution points.

Introduction - LP
So if we closely observe the optimal solutions lie on the intersection points or vertices. In LP terminology, the vertices are called extreme points So while finding out an optimal solution, we need not evaluate all the feasible solution points. We just have to consider only the feasible solutions that occur at extreme points.

Introduction - LP
So if we closely observe the optimal solutions lie on the intersection points or vertices. In LP terminology, the vertices are called extreme points So while finding out an optimal solution, we need not evaluate all the feasible solution points. We just have to consider only the feasible solutions that occur at extreme points. Here for Par Inc. problem, we have 5 extreme points and by evaluating these 5 points, we shall get the optimal solutions.

Example 1: A Maximization Problem

LP Formulation
Max 5x1 + 7x2

s.t.

x1

< 6

2x1 + 3x2 < 19 x1 + x2 < 8

x1, x2 > 0

Example 1: Graphical Solution

Constraint #1 Graphed
x2
8 7 6 5 4 3 2 1 1 2 3 4 5 6 7 8 9 10

x1 < 6

(6, 0) x1

Example 1: Graphical Solution

Constraint #2 Graphed
x2
8 7 6 5 4 3 2 1 1 2 3 4 5 6 7 8 9 10

(0, 6

1/3)

2x1 + 3x2 < 19 (9

1/2,

0)

x1

Example 1: Graphical Solution

Constraint #3 Graphed
x2
8 7 6 5 4 3 2 1 1 2 3 4 5 6 7 8 9 10

(0, 8) x1 + x2 < 8

(8, 0) x1

Example 1: Graphical Solution

Combined-Constraint Graph
x2
8 7 6 5 4 3 2 1 1 2 3 4 5 6 7 8 9 10

x1 + x2 < 8 x1 < 6

2x1 + 3x2 < 19 x1

Example 1: Graphical Solution

Feasible Solution Region
8 7 6 5 4 3 2 1 1 2 3 4 5 6 7 8 9 10

x2

Feasible Region x1

Example 1: Graphical Solution

Objective Function Line
8 7 6 5 4 3 2 1 1 2 3 4 5 6 7 8 9 10

x2

(0, 5)

Objective Function 5x1 + 7x2 = 35

(7, 0) x1

Example 1: Graphical Solution

Optimal Solution
8 7 6 5 4 3 2 1 1 2 3 4 5 6 7 8 9 10

x2

Objective Function 5x1 + 7x2 = 46 Optimal Solution (x1 = 5, x2 = 3)

x1

Prepare a graph of the feasible solutions for each of the constraints.

Summary of the Graphical Solution Procedure for Maximization Problems

Determine the feasible region that satisfies all the constraints simultaneously..

Draw an objective function line.

Move parallel objective function lines toward larger objective function values without entirely leaving the feasible region.

Any feasible solution on the objective function line with the largest value is an optimal solution.

Slack and Surplus Variables

A linear program in which all the variables are non-negative and all the constraints are equalities is said to be in standard form.

Standard form is attained by adding slack variables to "less than or equal to" constraints, and by subtracting surplus variables from "greater than or equal to" constraints.

Slack and surplus variables represent the difference between the left and right sides of the constraints.

Slack and surplus variables have objective function coefficients equal to 0.

Example 1
Standard Form Max 5x1 + 7x2 + 0s1 + 0s2 + 0s3

s.t.

x1 2x1 + 3x2 x1 + x2

+ s1

= + s2

6 = 19

+ s3 = 8

x1, x2 , s1 , s2 , s3 > 0

Extreme Points and the Optimal Solution

The corners or vertices of the feasible region are referred to as the extreme points. An optimal solution to an LP problem can be found at an extreme point of the feasible region. When looking for the optimal solution, you do not have to evaluate all feasible solution points. You have to consider only the extreme points of the feasible region.

Example 1: Graphical Solution

The Five Extreme Points
8 7 6 5 4 3 2 1

x2

4 Feasible Region 1
1 2 3 4 5 6

3 2
7 8 9 10

x1

Computer Solutions
Computer programs designed to solve LP problems are now widely available. Most large LP problems can be solved with just a few minutes of computer time. Small LP problems usually require only a few seconds. Linear programming solvers are now part of many spreadsheet packages, such as Microsoft Excel.

Example 1: Spreadsheet Solution

B D Partial A Spreadsheet ShowingCProblem Data 1 2 3 4 5 6 Constra ints #1 #2 #3 Obj.Func.Coeff. LHS Coe fficients X1 X2 1 0 2 3 1 1 5 7 RHS Va lues 6 19 8

Example 1: Spreadsheet Solution

B Partial A Spreadsheet ShowingCSolution 8 9 10 11 12 13 14 15 16 17 Optimal Decision Va ria ble Value s X1 X2 5.0 3.0 46.0 RHS Lim its 6 19 8 D

Max im ize d Objective Function Constra ints #1 #2 #3 Amount Use d 5 19 8

<= <= <=

Example 1: Spreadsheet Solution

Interpretation of Computer Output We see from the previous slide that:
Objective Function Value = 46 Decision Variable #1 (x1) = Decision Variable #2 (x2) = Slack in Constraint #1 Slack in Constraint #2 Slack in Constraint #3 = = = 5 3 1 (= 6 - 5) 0 (= 19 - 19) 0 (= 8 - 8)

Reduced Cost
The reduced cost for a decision variable whose value is 0 in the optimal solution is the amount the variable's objective function coefficient would have to improve (increase for maximization problems, decrease for minimization problems) before this variable could assume a positive value. The reduced cost for a decision variable with a positive value is 0.

Example 1: Spreadsheet Solution

Adjustable Cells Reduced Costs Final Reduced Cell $B$8 $C$8 Name Value X1 5.0 X2 3.0 Cost 0.0 0.0 Objective Coefficient 5 7 Allowable Increase Allowable Decrease 2 0.333333333 0.5 2

Constraints Final Shadow Constraint Allowable Allowable Cell Name Value Price R.H. Side Increase Decrease $B$13 #1 5 0 6 1E+30 1 $B$14 #2 19 2 19 5 1 $B$15 #3 8 1 8 0.333333333 1.666666667

Example 2: A Minimization Problem

LP Formulation Min s.t. 5x1 + 2x2

2x1 + 5x2 > 10 4x1 - x2 > 12 x1 + x2 > 4 x1, x2 > 0

Example 2: Graphical Solution

Graph the Constraints Constraint 1: When x1 = 0, then x2 = 2; when x2 = 0, then x1 = 5. Connect (5,0) and (0,2). The ">" side is above this line. Constraint 2: When x2 = 0, then x1 = 3. But setting x1 to 0 will yield x2 = -12, which is not on the graph. Thus, to get a second point on this line, set x1 to any number larger than 3 and solve for x2: when x1 = 5, then x2 = 8. Connect (3,0) and (5,8). The ">" side is to the right. Constraint 3: When x1 = 0, then x2 = 4; when x2 = 0, then x1 = 4. Connect (4,0) and (0,4). The ">" side is above this line.

Example 2: Graphical Solution

x2 Constraints Graphed Feasible Region 4x1 - x2 > 12
5 4 3 2 1

x1 + x2 > 4 2x1 + 5x2 > 10

x1
1 2 3 4 5 6

Example 2: Graphical Solution

Graph the Objective Function Set the objective function equal to an arbitrary constant (say 20) and graph it. For 5x1 + 2x2 = 20, when x1 = 0, then x2 = 10; when x2= 0, then x1 = 4. Connect (4,0) and (0,10). Move the Objective Function Line Toward Optimality Move it in the direction which lowers its value (down), since we are minimizing, until it touches the last point of the feasible region, determined by the last two constraints.

Example 2: Graphical Solution

Objective Function Graphed
x2 Min z = 5x1 + 2x2 4x1 - x2 > 12
5 4 3 2 1

x1 + x2 > 4 2x1 + 5x2 > 10

x1
1 2 3 4 5 6

Example 2: Graphical Solution

Solve for the Extreme Point at the Intersection of the Two Binding Constraints 4x1 - x2 = 12 x1+ x2 = 4 Adding these two equations gives: 5x1 = 16 or x1 = 16/5. Substituting this into x1 + x2 = 4 gives: x2 = 4/5

Example 2: Graphical Solution

Solve for the Optimal Value of the Objective Function Solve for z = 5x1 + 2x2 = 5(16/5) + 2(4/5) = 88/5. Thus the optimal solution is
x1 = 16/5; x2 = 4/5; z = 88/5

Example 2: Graphical Solution

x2 Optimal Solution
5 4 3 2 1 1 2 3 4 5 6

Min z = 5x1 + 2x2 4x1 - x2 > 12 x1 + x2 > 4 2x1 + 5x2 > 10 Optimal: x1 = 16/5 x2 = 4/5 x1

Example 2: Spreadsheet Solution

A B D Partial Spreadsheet Showing C Problem Data 1 2 Constra ints 3 #1 4 #2 5 #3 6 Obj.Func.Coe ff. LHS Coefficients X1 X2 2 5 4 -1 1 1 5 2 RHS 10 12 4

Example 2: Spreadsheet Solution

A B Partial Spreadsheet Showing C Formulas D 9 Decision Variables 10 X1 X2 11 Dec.Var.Values 12 13 Minimized Objective Function =B6*B11+C6*C11 14 15 Constraints Amount Used Amount Avail. 16 #1 >= =D3 =B3*$B$11+C3*$C$11 17 #2 >= =D4 =B4*$B$11+C4*$C$11 18 #3 >= =D5 =B5*$B$11+C5*$C$11

Example 2: Spreadsheet Solution

A B Partial Spreadsheet Showing C Solution 9 Decision Variables 10 X1 X2 11 Dec.Var.Values 3.20 0.800 12 13 Minimized Objective Function 17.600 14 15 Constraints Amount Used 16 #1 10.4 >= 17 #2 12 >= 18 #3 4 >= D

Amount Avail. 10 12 4

Feasible Region
The feasible region for a two-variable linear programming problem can be nonexistent, a single point, a line, a polygon, or an unbounded area. Any linear program falls in one of three categories: is infeasible has a unique optimal solution or alternate optimal solutions has an objective function that can be increased without bound A feasible region may be unbounded and yet there may be optimal solutions. This is common in minimization problems and is possible in maximization problems.

Special Cases
Alternative Optimal Solutions In the graphical method, if the objective function line is parallel to a boundary constraint in the direction of optimization, there are alternate optimal solutions, with all points on this line segment being optimal. Infeasibility A linear program which is overconstrained so that no point satisfies all the constraints is said to be infeasible. Unboundedness (See example on upcoming slide.)

Example: Infeasible Problem

Solve graphically for the optimal solution: Max s.t. 2x1 + 6x2

4x1 + 3x2 < 12 2x1 + x2 > 8 x1, x2 > 0

Example: Infeasible Problem

There are no points that satisfy both constraints,x2 hence this problem has no feasible region, and no optimal solution.
8

2x1 + x2 > 8

4x1 + 3x2 < 12

3 4

x1

Example: Unbounded Problem

Solve graphically for the optimal solution: Max
s.t. x1 + x2 > 5 3x1 + x2 > 8

3x1 + 4x2

x1, x2 > 0

Example: Unbounded Problem

The feasible region is unbounded and the objective function line can be moved parallel to itself without bound so that z can be increased infinitely. x2 3x1 + x2 > 8 8 5 Max 3x1 + 4x2 x1 + x2 > 5

2.67

x1

Introduction - LP
Minimisation problem: M&D Chemicals produces 2 products A & B. Based on analysis, M&D management specified the following - Min. production of both A & B should be 350 gallons - One major customers order of 125 gallons of product A must be produced - Product A requires 2 hours of processing time and product B requires 1 hour per gallon - Total available hours in the next month are 600 - Production costs are $2 and $3 per gallon of product A B respectively. Find the minimum cost production schedule

Introduction - LP
Ans.: LP model for M&D problem can be written as follows Minimise 2A+3B subjected to A 125 demand for product A A + B 350 total production 2A + B 600 A, B 0 Draw the constraint lines on the graph using Product A on x axis and product B on y axis. Find out the feasible region. Find the extreme points i.e. vertices Find out the production cost using the formula 2A + 3B for various extreme points and the lowest is the right answer. processing time

Introduction LP
Surplus variable: In case of M&D problem, the solution is A = 250 and B = 100 As we see, we have produced 125 gallons (250-125) excess as opposed to the min. required quantity. This excess production in LP is called Surplus. So in LP any excess quantity corresponding to is referred to as surplus. So for a constraint with , surplus variable is subtracted.

Introduction - LP
So with the surplus variables, the LP for M&D looks like Min 2A + 3B + 0S1 + 0S2 + 0S3 subjected to A - S1 = 125 A + B - S2 = 350 2A + B + S3 = 600 A, B, S1, S2, S3 0

Introduction - LP
Solve the M&D problem using MS Excel