CHAPTER 2 Linear Programming: Graphical Method Learning Objectives At the end of this chapter, the student should be able to: 1. explain linear programming; 2. identify the parts of linear programming; 3. state the assumptions in linear programming; 4, determine the feasible region; and 5. compute for the optimal solution using the graphical method. OVERVIEW Linear programming is a quantitative technique used in properly allocating the resources of a business to maximize its profit or minimize cost. The term linear signifies the existence of a proportional relationship between variables, which can either be direct or inverse. A directly Proportional relationship exists when an increase in one variable corresponds to an increase in another variable. An inversely proportional relationship exists when an increase in one variable results in a decrease in another variable or vice versa.1 10 Muwcewent Science: A Questa Arron To Don MAKING Programming refers to the use o' a mathematical model when developing an equation to | solve a problem. The term is not related to computer programming. ‘The following are the four properties of linear programming problems: 1. Presence of objective function. The objective of a linear programming problem is either to maximize profit or minimize cost. The term profit is the difference between > the revenue and the variable cost. The term cost simply refers to the variable cost. The fixed cost is not included because itis not linearly related to the revenue. The fixed cost | remains constant in relation to the changes in revenue and production, The objective of | a linear programming problem must be expressed in an equality. 2. Presence of constraints. There are constraints that restrict or limit the attainment of an_ objective. The limitations may be in the form of plant capacity, availability of materials, direct labor or personnel, or time. The limitations or constraints should not be violated in determining the optimum solution. The constraints or limitations must be expressed in equalities or inequalities. 3. Availability of alternatives. Since there are two or more variables used when developing a linear programming problem, there are different alternative courses of action to choose from. The alternative courses of action may favor the selection of one variable over another or it may involve the mixture of two or more variables, 4, Linear mathematical relationship. This property means that the relationship of the variables in linear programming is considered linear and can be expressed in equalities or inequalities. The objectives, functions, and constraints must be in their first degree level (eg, 3A + 4B = 25). ASSUMPTIONS OF LINEAR PROGRAMMING 1. Certainty assumption. There is certainty that the objective function and the constraints will not change during the pericd of the study. The mathematical equation that has been developed should remain the same as well. 2. Proportionality assumption. Proportionality exists among the variables in the objective function and constraints. For example, if one unit of product requires 4 kilos of raw materials, then producing five units requires 20 kilos of raw materials. 3. Additivity assumption, There is additivity in the mathematical model. This assumption means that the sum of all activities is equal to the sum of the different individual activities. For example, if one product contzibutes P6 to the profit of a business and another product contributes a profit of P3, then the total profit for the two products is P9. 4, Divisibility assumption. The optimum solution may take the form of a fraction, not necessarily an integer or a whole number. However, an integer programming problem will exist when a fractional form results in manufacturing a product. For example, manufacturing three and three-fourths chairs is not possible.(Conpren2 | Linens Proceaununs: Geamicar Memon 11 5. Non-negativity assumption. The value of the decision variables should be positive or zero. This assumption is also known as logical constraint since a company cannot produce a negative number of products. PARTS OF A LINEAR PROGRAMMING MODEL A linear programming model has two main parts, namely, the objective function and the constraints. The objective function denotes the maximization of profit or minimization of cost. tis usually introduced by the words “maximize” or “minimize.” The constraint is introduced by the words “subject to.” There are two types of constraints, namely, the explicit constraint and the implicit constraint. Explicit constraints are limitations that are expressed or stated in a problem. Implicit constraints are implied limitations or non-negativity assumptions. These constraints should not be violated. METHODS OF SOLVING A LINEAR PROGRAMMING PROBLEM A linear programming problem may be solved by using either the graphical method or the simplex method. GRAPHICAL METHOD The graphical method utilizes a two-dimensional graph to solve linear programming problems. This method is used if the linear programming model consists of no more than two variables. ‘The following are the steps involved in the graphical method: 1. Use letters to represent the unknowns or variables in the problem. . If necessary, prepare a simple matrix to facilitate the formulation of the constraints. Formulate the objective function and the constraints in equalities and inequalities . Graph the constraints. Ifthe lines intersect, solve for the coordinates of the intersection point. ;. Determine the feasible region. . Determine the optimum solution. . Formulate the decision based on the highest profit or lowest cost: The feasible region contains all the points in the graph where the optimum solution can be found. It is also the region where the constraints are not violated. There are two methods for finding the optimal solution, namely, the isoprofit or isocost line method and the corner point method. In this book, the corner point method will be used.EARS 12. Mrwecoonr Sconce: A Questa Arron 10 Deoson Mass Bins ee eee eee Maximization of Profit (Graphical Method) Jenny Furniture Shop produces two types of tables: an elegant table and an ordinary table. The table, whether elegant or ordinary, passes through two departments, namely, the assembly department and finishing department. Each elegant table requires 8 hours of assembly work and 4 hours of finishing tasks. Each ordinary table requires 6 hours in the assembly department and 2 hours in the finishing department. Every production period, the shop has 480 hours of assembly time and 200 hours of finishing time. The elegant table contributes #700 to the company’s profit, while the ordinary table has a profit of 500. Required. Using the graphical method of linear programming, determine the number of units needed to be produced to maximize profits. ‘Answer. The constraint in this case is the availability of working hours in the assembly and finishing departments. Applying the steps, the number of units to be produced to maximize profits is determined as follows: Step 71 4 Use letters to represent the unknown variables in the problem. Let x = number of elegant tables to be produced y = number of ordinary tables to be produced Step2 If necessary, prepare a simple matrix to facilitate the formulation of the constraints. Hours Required per Unit x y Available Hours Assembly department 8 6 480 Finishing department 4 2 200 Profit 700 500 Step 3 Formulate the objective function and the constraints in equalities and inequalities. As the number of hours in the assembly and finishing departments may not exceed 480 and 200 hours, respectively, the < shall be used to express the constraints. Max profit: P = P700x + P500y Subjectto: 8x + 6y $480 4x + 2y $200 xy20Cure? | Ln Proceaainc: Gonscn Meno 13 Step 41 Graph the constraints. In order to do this, the inequality must first be converted to an equality to determine the intercept. For the first constraint: 8x + 6y $480, the equation will be: ox + 6y = 480 The intercept of this equation is determined as follows: ‘fx= 0, then y is 80, which is computed Ify = 0, then x is 60, which is computed as follows: as follows: 0 + by = 480 & + 0 = 480 y = 80 x = 60 For the second constraint: 4x + 2y < 200, the equation will be: 4x + 2y = 200 The intercept of this equation is determined as follows: If x=0, then y is 100, which is computed Ify = 0, then x is 50, which is computed as follows: as follows: o+ 4x + 0 = 200 x = 50 In summary, the equation and the intercepts are: For equation 8x + 6y = 480, the intercepts or vertices are (0, 80) and (60, 0). For equation 4x + 2y = 200, the intercepts or vertices are (0, 100) and (60, 0).EZR 14 MnwcenensScance: A Queenie Armen 10 Dow Mans Figure 2.1 presents the graph of the coordinates. © 10 29 30 40 $0 60 70 80 90 100 Figure 2.1 The graph shows that the two lines intersect. Hence, there is a need to solve for the coordinates of the point of intersection. By elimination process, the coordinates are determined as follows: & + by = 480 x + Multiply the second equation by 2, then deduct the second equation from the first to eliminate x. Afterward, solve for the value of y as follows: & + by = 480 & + 4y = 400 Bs Substitute the value of y, either to the first or second equation, to solve for the value of x. Substituting the value of y to the first equation, the value of x is determined as follows: & + by 480 8 + 640) = 480 x = 390 Thus, the coordinates of the point of intersection are (30, 40),wre? | Lneat Proctananc: Gewmacn Meri00 15 5 Determine the feasible region. There are five coordinates: (0, 100), (0, 80), (60, 0), (50, 0), and 40). At this time, one can hardly identify where the feasible region is by simply looking at graph. Will the feasible region be bounded by the following vertices? 1. (0, 100), (0, 80), and (30, 40) 2. (60, 0), (50, 0), and (30, 40) 3. (0, 100), (80, 40), and (60, 0) 4, (0, 80), G0, 40), and (60, 0) 5. (0, 80), (30, 40), and (50, 0) ‘It is highly emphasized that the feasible region is the area where the constraints are ied. Thus, the coordinates at the vertices, including that of the intersection point, shall be ested individually against all the constraints. To test these, the values of the coordinates are mply substituted to the constraints. If the coordinates violate one of the constraints, such coordinates will not form part of the feasible region. The coordinates in the graph are as follows: (0,100), 0, 80), (60, 0) (60,0), and (30, 40) 480 200 Theconstraints are: 8x + ~ 6y 4x + oy < < Coordinates (0, 100) First constraint 8x + 6y << 480 8(0) + 6100) < 480 600 < 480 ‘As computed, the coordinates (0, 100) violate the first constraint since 600 is not lower than 480. In such a case, there is no need for the coordinates (0, 100) to be tested against the second constraint. Hence, the coordinates (0, 100) will not form part of the feasible region. Coordinates (0,80) Firstconstraint - & + 6y < 480 80) + 680) < 480 480 s 480 The coordinates (0, 80) do not violate the first constraint since 480 is equal to 480. These coordinates are then tested against the second constraint as follows: Second constraint 4x + 2y $200 4) + 20) < 200 160 «=< 200ZZ 16 Wowooven Some: A Quoename Arma 10 Deas Mans The second limitation is not violated by the coordinates (0, 80) as well. As the coordinates (80) do not violate any of the two constraints, they shall form part of the feasible region. Coordinates (60, 0) | First constraint 8x + 6y << 480 860) + 60) < 480 4380s (480 The coordinates (60, 0) do not viclate the first constraint since 480 is equal to 480. However, they shall be tested against the second constraint as follows: | Second constraint: 4x + 2y < 200 | 4(0) + 20) < 200 240s 200 ‘This time, the second constraint is not satisfied. Hence, the coordinates (60, 0) will not form part of the feasible region. Coordinates (50, 0) First constreint 8x + fy =< 480 860) + 60) < 480 400 < 480 The first limitation is not violated by the coordinates (50, 0) since 400 is lower than 480. It shall, however, be tested against the second limitation. Second constraint: 4x + 2y < 200 460) + 20) < 200 200 < 200 The second limitation is not violated by the coordinates (60, 0). Thus, these coordinates shall be included as part of the feasible region. Coordinates (30, 40) : First constraint: 8 + 6y — < 480 8@0) + 640) < 480 480 < 480 The first constraint is not violated. The coordinates (30, 40) shalll be tested against the second constraint as follows: Second constraint: 4x + 2y < 200 4@0) + 240) < 200 200 < 200(Cousre 2 | Lness Paocaamunc: Gearicn. Merioo 17 Since the coordinates (30, 40) do not violate any of the constraints, they shall be included in the feasible region. At this point, the feasible region is bounded by the following vertices: (0, 80), 60, 0), and 30, 40). ‘The shaded portion in Figure 2.2 is the feasible region. Any coordinate located within such a region does not violate any of the constraints. S15 S08 208 0 10 20 30 40 50 60 70 80 90 100 Figure 22 Step 6 Determine the optimum solution. The two methods for finding the optimum solution are the isoprofit or isocost line solution method and corner point solution method. In the first method, a series of parallel lines will be drawn in the feasible region until the highest profit ‘or lowest cost is found. The second method is based on the mathematical theory behind linear programming in which the optimum solution lies at one of the corner points of the feasible region. This book illustrates the second method. Substituting the coordinates (0, 80), (50, 0), and (30, 40) in the vertices comprising the feasible region, the maximum profit is determined as follows: Vertices Max profit P700x P500y (0,80) 700(0) 50080) 60,0) 700660) 5000) 60, 40) 70080) 500(40) Step7 Formulate the decision based on the highest profit or lowest cost. The optimum sol therefore, is to produce 30 units of elegant tables and 40 units of ordinary tables to realize a profit of P41,000.00.z= 48 Manncevenr Scunce: A Quatre ArROHCH To DEcMioN: MAN peg eco ie ene Minimization of Cost (Graphical Method) To operate the newly imported state-of-the-art machinery recently purchased by ABC Company, two types of fuels are needec, namely, Fuel A and Fuel B, The machine specifications and requirements show that for every one hour of production time, each gallon of Fuel A emits 6 units of chemical gas and generates a mechanical energy equivalent to 8 horsepower. This costs P320. In a similar manner, each gallon of Fuel B emits 10 units of chemical gas and generates 8 horsepower of mechanical energy. This costs P300. The maximum number of units of such type of chemical gas allowed by the government to be emitted by such type of machinery is 30 units. The operating activities of the company require at least 32 horsepower of mechanical energy per hour. Required. Using the graphical method, determine the number of gallons of Fuel A and Fuel B that are needed to operate the machine at the lowest possible cost. Answer. There are two constraints in this case, namely, the maximum number of units of chemical gas allowed to be emitted and the required horsepower of mechanical energy. The steps involved are as follows: Step7 Use letters to represent the unknown variables in the problem. Let x = number of gallons of Fuel A y = number of gallons of Fuel B Sep? If necessary, prepare a simple matrix to facilitate the formulation of the constraints. The matrix appears as follows: Units a ane Rischiik ee x eee 8 7320.00 10 8 7300.00 30 2 Step 3 Formulate the objective function and the constraints in equalities and inequalities. The problem used the words “maximum” and “at least.” The former indicates that the required units or quantities should not be more than the given number, and the latter means that the required‘Cuore 2 | Lnean Proceamuns: Ganracas Merioo 19 units should not be less than the given number. Hence, < will be used for the former and 2 will be used for the latter. Min cost: C= 320x + P300y Subject to: 6x + 10y <30 8x + 8y 232 xy20 ‘Step 41L_ Graph the constraints. The inequality must first be converted to an equality, and the intercept should be determined before graphing the constraints. For the first constraint: 6x + 10y < 30, the equation will be: & + Wy = 30 The intercept of this equation is determined as follows: Ifx= 0, then y is 3, which is computed If y = 0, then x is 5, which is computed as follows: as follows: 0 + Wy = 30 B38. For the second constraint: 8x + 8y 2 32, the equation will be: & + By = 32 The intercept of this equation is determined as follows: Ifx=0, then y is 4, which is computed Ify = 0, then x is 4, which is computed as follows: as follows: 0 & + 0 = x 7 In summary, the equation and the intercepts are: For equation 6x + 10y = 30, the intercepts or vertices are (0,3), (5,0) For equation 8x + 8y = 32, the intercepts or vertices are (0,4), (4, 0)BESS 20 Movaners Soaice A Quemane Arron 1 Deco Mins Figure 2.3 presents the graph of the coordinates, Figure 2.3 The two lines intersect; hence, there is a need to determine the coordinates of the point of intersection. By elimination process, the coordinates are determined as follows: & + by 30 & + By = 32 To eliminate x, multiply the first equation by 8 and the second equation by 6. Then, deduct the second equation as follows: 4x + S0y = 240 4x + 48y = 192 5 Sy = 48 y 15 Substitute the value of y to the first equation to determine the value of x as follows: & + Wy = 30 ox + 1005) = 30 x = 25 The coordinates of the point of intersection, then, are (2.5, 15).(Civerte2 | Linens Procaaannc: Gearwca Mero 21 Determine the feasible region. The coordinates of the vertices, including the intersection it, are (0, 4), (0,3), (2.5, 1.5), (4, 0), and (5, 0), These coordinates will be tested to identify which them will satisfy the two constraints. ‘Coordinates (0,4) Firstconstraint x. + “Wy << 30 60) + 104) < 30 40 << 30 The coordinates (0, 4) violate the first constraint; hence, they do not have to be tested against the second constraint anymore. These coordinates shall not form part of the feasible region. First constraint: & + Idy 60) + 108) 30 ‘The coordinates (0, 3) satisfy the first constraint. Hence, they shall be tested against the constraint as follows: Second constraint: 8x + 8y 2 32 80) + 8G) = 32 m4 2° R This time, the coordinates (0, 3) violate the second constraint because 24 is not greater 32, Hence, they will not form part of the feasible region. tes (2.5, 1.5) First constraint: 6x. + — dy 62.5) + 100.5) 15K + cet 30BB 22. Maxacowenr Scence: A Quaroanive ArpRoAci 10 Deision- Manne The coordinates (2.5, 1.5) do not violate the first constraint, Hence, they shall be tested against the second constraint as follows: Second constraint: &x + sy 2 32 a5) + 85) 2 32 0 + 2 2 @ 2 2m The coordinates (2.5, 1.5) satisfy the two constraints. Hence, they will form part of the feasible region. Coordinates (4, 0) First constraint & + Wy s 30 64) + 100) < 30 4 < 30 The coordinates (4, 0) satisfy the first constraint. Hence, they shall be tested against the second constraint as follows: Second constraint: 8x + wy 2.3 84) + 80) = 32 2 > 2 The second constraint is not violated by the coordinates (4, 0). Thus, they will form part of the feasible region. Coordinates (5, 0) First constraint & + Wy s< 30 66) + 100) < 30 30 s 30 The coordinates (5, 0) do not violate the first constraint because 30 is equal to 30. They shall be tested against the second constraint as follows: Second constraint: 8 + 8y 2 32 86) + 80) > 32 40 2 2 The coordinates (5, 0) do not violate the second constraint as well because 40 is higher than 30, These coordinates will also form part of the feasible region. At this time, the coordinates of the vertices comprising the feasible region are (2.5, 1.5), (4, 0), and (5, 0), The shaded portion of Figure 2.4 shows the feasible region.CChaprer2 | Lveat Paoceannanc: Gearaca, Memwoo 23 Step 61 Determine the optimum solution. By substituting in the objective function the coordinates Of the vertices located at the feasible region, the computation will appear as follows: Vertices Min cost 7320x + = P300y 25,15) = 32025) 300(1.5) 1,250.00 + 40 3204) + += 3000) 1,280.00 + 69 32066) 300(0) ‘71,600.00 Step7 Formulate the decision based on the highest profit or lowest cost. Since the objective function is to minimize the cost, the decision, then, is to use 2.5 gallons of Fuel A and 15 gallons of Fuel B for the optimum cost of 1,250.00. SPECIAL CASES IN LINEAR PROGRAMMING There are four special cases that may arise when solving linear programming problems using the graphical method, namely, infeasibility, unboundedness, redundancy, and alternate optimal solution.aA 24 Manwoevent Soence: A QuantTArne APPROACH TO DEcsON- MAINS Infeasibility ‘The problem of infeasibility occurs when there are no solutions to a linear programming problem that could meet the requirements of the constraints. This happens when the constraints are conflicting, and no feasible region can be determined. ‘To illustrate, consider the following explicit constraints: x + 4 < 2 4x + 2y 6 x 8 vw Figure 2.5 shows the graph of the constraints. Qesiea 34-5. 6.7 89:10 Figure 2.5 ‘There is no feasible region because of the third constraint, which is a conflicting constraint. Unboundedness Unboundedness occurs when the maximization problem can have infinitely large values without violating the-requirements of the constraints. This happens when the linear programming constraints are wrongly formulated, thus resulting in one or more missing constraints.‘Copter 2 | Lncas Proceamunc: Gexmica. Merion 25 To illustrate, consider the following explicit constraints: x + 4 2 0 2x > 10 y < 6 The graph will appear as follows: 10 ays 16 ee 10 Feasible region (Pee We et eee a Figure 2.6 Figure 2.6 shows that the feasible region that extends infinitely to the right will result in an yunded profit, which is not possible because the resources of a business have limitations. Redundancy . Redundancy happens when there is a redundant constraint, such as a requirement or tation, that will not affect the feasible region. This indicates that the feasible region can still ‘be determined even if the redundant constraint is removed from the model. To illustrate, consider the following explicit constraints: m+ y os 3 Sb ys Se s x 28BA 26 Maxacowerr Scrat: A Quasar APPROACH 70 DecsioN-MAKING The graph will appear as follows Figure 27 In Figure 27, the constraint x < 28 is considered redundant since it does not have an effect or bearing on the feasible region. Alternate Optimal Solution Ina problem with an alternate optimal solution, two or more optimal solutions exist with the same level of profit regardless of the selected combination. This happens when the slope of the isoprofit or isocost is perfectly parallel to one of the constraints in the problem. ‘To illustrate, consider the following linear programming model: Max profit: P= 6x + 4y Subjectto: Sx +2y S12 xs3‘Chuerer 2 | Linens Proceanaanc: Gearuca Memoo 27 The graph will appear as follows: Figure 2.8 The coordinates of the intersection are (3, 1.5). In Figure 2.8, the optimal solution is found points A and B whose coordinates are (0, 6) and 3, 15), respectively. Any combination xand y along this line will produce the same level of profit. To illustrate, consider two coordinates only: (0, 6) and @, 15). The objective function is = 6x + 4y. Substituting the values, the profit is determined as follows: @6) P= & + gy ee): = 4-4 & + dy 6G) + 4(15) 4