Instituto Politéenico de Tomar cola Superior de Tecnologia (0 de Engenharia Electrotéenica to © Gestio de Projectos 2008-2009Planeamento e Gestao de Projectos Capitulo 1 Programagio Linear e Métodos de Previ Programagao Linear: Coneceito de optimizagao: Exemplos de motivagio no contexto empresaria COptimizagao sem restrigdes e com restrigdes Variaveis de decisto; Formullagio ¢ propriedades bisicas do problema; Resolucdo geifica de Problemas de Programagao Linear: Algoritmo Simplex: Dualidade e andlise de sensibilidade, Métodos de Previsdo Média movels ‘Alisamento exponencial com tendéncia e sazonatidade, Regressio linear e nfo linear erro da previsio:Planeamento ¢ Gestio de Projectos Exemplo | A empresa Giapetto fabrica dois tipos de brinquedos de madeira ~ soldados e comboios. Um soldado é vendido por 27€ e usa 10 € de materia prima (MP). Cada soldado fabricado tem 14€ de gastos erais de fabrico (GGF) e mao de obra directa (MOD) variével. Um comboio ¢ vendido por 216, usa 9€ de MP ¢ tem 10€ de GGF e MOD variavel © fabrico de soldados e comboios de madeira tem 2 tipos de trabalho - Carpintaria @ Acabamento, Um soldado requer 2h de acabamento e 1h de carpinaria. Um comboio requer 1h de acabamento © Th de carpintaria Ent ceda semana a empress tm disponivel oda a MP recess, mas apenas tem 100h de acabamento {80h de carpintaria ‘A demanda de comboios il soldados. ‘A-empresa pretende maximizar o Iucro semanal (recitas ~ custo). Formato o modelo matemtico para a maximizagao do luero semana nitada enguanto que em cada semana so necessirios no maximo 40 Exemplo 2, A empresa Dorian Auto Fabrica automéveis e autocarros de luxo. A companhia sabe que os clientes mais provaveis so homens ¢ mulheres de altos rendimentos. Para aleangar este grupo de clientes @ empresa decidiu investir em publicidade televisiva, comprando pacotes de 1 minuto de aniincios em dois tipos de programas: especticulos de comédia e jogos de futebol. Cada anuncio de comédia € visto por 7 milhies ‘de mulheres com altos rendimentos e 2 milhes de homens com altos rendimentos. Cada anuncio de futebol é visto por 2 miles de mulheres com altos rendimentos e 12 milhGes de homens com altos rendimentos. Um minuto de anuncio de comédia custa 50,000€ ¢ um minuto de anuncio de futebol custa 100.0006. A empresa queria que os comerciais fossem vistos por pelo menos 28 mihdes de mulheres com altos rendimentos, ¢ 24 milhdes de homens de altos rendimentos. Use a Programagio Linear para determinar como & que a empresa deve alcancar os seus objectivos com um custo minimo. Exemplo 3. Uma industria automével fabriea automéveis ¢ autocarros. Cada veiculo tem que ser processado na zona de pintura e na zona de montagem. Se a zona de pintura estiver apenas a pinlar autocarros, podem ser pintados 40 por dia, Se a zona de pintura estiver apenas a pintar automéveis, podem ser pintados 60 por dia, Se a zona de montagem estiver a montar autocarros, podem ser processados 50 por dia, se estiver a ‘montar automéveis, podem ser processados também 50 por dia. Cada autocarro contribui com 300€ para o Tero e cada automével contribui com 2006, Use a programagao linear para determinar a produgio dideia {que maximiza os lucros da companhia, Exemplo 4 Suponha que a empresa do exemplo 3 quer produzir pelo menos 30 autocartos 20 automéveis. Fncontre a solugo dptima do novo problema Exempla 5, Resolva graficamente o seguinte problema de programagdo linear max st. -m) SI 2x, +3 26 x %y 20Planeamento e Gestdo de Projectos Exemplo 6 Considere uma dieta constituida apenas por 4 alimentos: Al, A2, A3e A4. Cada alimento Al, A2, A3 © A4 costa 0.5€/unidade, 0.20€/unidade, 0,30€/unidade e 0,80€/unidade, respectivamente. Diariamente, devem ser ingeridas pelo menos 500 calorias, 6g de Vitamina C, 10g de Vitamina E e Se de Vitamina B. O conteido nutricional por unidade de cada alimento est apresentado ra tabela em baixo. Formule o modelo de programacao linear que satisfaga as necessidades nutrcionais com o custo minimo, Catorias —-Vitamina Cg] —-Vitamina E[g]_Vitarnina Ble} palma 400) 7 — 2 re 200 2 2 4 a3 150 0 4 ' Ad 500 0 4 5 Exemplo 7. Considere um posto dos correios que necessita de um nimero diferente de empregados em diferentes dias dda semana (Tabela 1). A lei impOe que cada empregado tem que trabalhar einco dias consecutivos seguidos de 2 dias de folga. Por exemplo um empregado que trabalha de terga a sibado, tem que tirar folga ‘no domingo e na segunda. Formule o problema de Programa Linear para que o posto dos correios possa. ‘minimizar o n° de empregados contratados. [N? de empregados necessirios Segunda "7 Terga 3 Quarta 15 Quinta 9 Sexta 4 Siibado 16 Domingo "1 Vabela 1 Exemplo 8 A-empresa Sunco produz trés tipos de gasolina (1, 2 e 3), Cada tipo ¢ obtido pela mistura de tés tipos de crude (1, 2&3). O prego de vend do barril de gasolina e o prego de compra do crude so dados na tabela 2. A-empresa pode comprar até 5000 barris de cada tipo de crude diariamente, (Os tréstipos de gasolina diferem na sua taxa de octanas e no surfactante, O crude misturado para formar a gasolina | tem que ter uma taxa de octanas de pelo menos 10 © no maximo 1% de surfactante. O crude rmisturado para obter a gasolina 2 tem que ter uma taxa de octanas de pelo menos 8 € no maximo 26 de surfactante, O crude misturado para obter a gasolina 3 tem que ter uma taxa de octanas de pelo menos 6 & no maximo 1% de surfactante, Os custos de transformar um baril de petréleo em um de gasolina sto de 4, € a refinaria pode produzir até [4,000 barris de gasolina por dia, Os clientes requerem diariamente 3000 barris da gasolina 1, 2000 barris da gasolina 2 © 1000 bartis de sgasolina 3 Cada € gasto diariamente em publicidade num determinado tipo de gasolina, aumenta a demanda automaticamente em 10 bartis. Formule 0 problema de P.L. de modo a que a empresa maximize o luero (Lucro = Receitas - Custos) Prego de venda por barr 70€ ne S0€ Gasolina 1 Gasolina 2 Gasolina 3Planeamento e Gestiio de Projectos Tabela2 | Preco de compra porbarril Taxa de Oetanas __Conteio de surfactant Gradet 45 2 o Crude 2 ase 6 20% Crude 3 25€ 8 30% ‘Tabela3 Exemplo 9. A empresa Rylon fabrica os perfumes Pi e P2. A matéria-prima necesséria para fabricar cada tipo de perfume € comprada a 3€ por libra. O processamento de Ib de matéria-prima necessita de 1h de trabalho de laboratério. Cada libra de matéria-prima da para fazer 3 ongas de perfume P2 Normal e 4 ongas de perfume PI Normal. O perfume P2 Normal pode ser vendido a 76/onga eo Pl Normal a 6€/onga. A ‘empresa pode também produzir perfumes de luxo a partir do perfume Normal, 0 P2 de Luxo € vendido a 8€/onga © 0 PI de Luxo a 14€/onga, Uma onga de P2 de Luxo é obtida através de uma onga de P2 Normal com 3h de tempo de laboratdrio ¢ 4€ de custos de processamento. Cada onga de PI Normal necessita de 2h de laboratério e 4€ de custos de processamento para se transformar em PI de Luxo, A empresa tem disponiveis 6000h de laboratério anuais © pode comprar até 4000Ib de matéria-prima, Formule o problema de programagio linar que maximize os lueros, Assuma que os custos das horas de laboratérios sto custos fixes. Exemplo 10. ‘A empresa Saileo tem que determinar quantos barcos devem ser produzidos durante 0s préximos 4 twimestres. A demanda durante cada um dos trimestres Trimesire [tines 2 4 No principio do 1° trimestre a empresa tem no inventirio 10 bareos. No prinefpio de cada timestre a empresa tem que decidir quantos barcos vai produzit. Para simplificar, considera-se que os barcos fabricados durante o trimestre sfo usados para colmatar a respectiva demanda. Durante cada trimestre a Saileo pode produzir até 40 barcos utilizando trabalho normal ao custo de 400€ por barco ou utilizando horas extra com o custo final de 450€ por barco. No final de cada wrimestre (apés a produgdo e demand satisteita) & imputada 20€ por barco em custos de posse © armazenamento, Determine a calendarizagio da produgio que minimize a soma dos custos de jentirio e de produgo durante os prOximos quatro trimestres,Planeamento e Gestio de Projectos Capitulo 3 Exereicio 2 da pagina 76 Suponha que um posto dos correios pode forcar os seus empregados a trabalhar un dia extra em cada semana, Por exemplo, um empregado cujo horirio regular seja de segunda a sexta feira pode também trabalhar ao sibado, Cada empregado & pago a SO€ por cada dia do horario regular, e 62€ por dia extra (se existic), Formule o problema de programagao linear que permita o posto dos correios minimizar os custos com o pessoal Capitulo 4 Exercicio 5 da pagina 76 iariamente, 0s empregados de um departamento da policia trabalham 2 turos de 6 horas, com as seguintes necessidades de pessoal: WEimpregados 15 Os empregados que trabalham dois turmos consecutives so pagos 126hhora e aqueles que no sto consecutivos sto pagos a 18€/hora, Formule 6 problema de programago linear usado para minimizar os ceustos com 0 pessoal Capitulo 5 Exereicio 2 da pagina 91 A empresa de sumo Sumolis vende sacos de rede com laranjas e pacotes de sumo de laranja. As eategorias dda laranjas comercializadas tem categorias de 1 (fraca) a 10 (excelente). A Sumolis tem nos seus armazéns 100.000 kg de laranjas de categoria 9 © 120.000 kg de categoria 6. A qualidade media das laranjas vendidas em saco tem que ser pelo menos de 7, € a qualidade media das laranjas usadas para sumo tem 4que ser pelo menos 8. Cada ky de laranjas que sto usadas para sumo, dio 1,5€ de receitas e incorrem 1,05€ de custos (MOD, armazenamento, etc.). Cada kg de laranjas de saco € vendido a 0,50€ e custos variiveis de 0,20€. Formule o problema de programagio linear que maximize o lucto, Capitulo 6 Exercicio 5 da pagina 91 Uma empresa petrolifera tem 5.000 barris de petrSleo | € 10,000 de petréleo 2. A companhia vende dois produos: gasolina e gasolen de aguecimento, Ambos os produtos sia produzidas combinando o petréleo | © 2. Onivel de qualidade do petréleo 1 é de 10 e 0 pettileo 2 € 5. A gasolina tem que ter um nivel médio de qualidade de pelo menos 8 « 0 gasoleo de 6, A demanda destes produtos tem que ser criada apenas pela publicidade. Cada euro gasto em publicidade da gasolina cria uma demanda de 5 barris e cada euro gasto em publicidade do gasoleo cria uma demanda de 10 barris. A gasolina & vendida a 25€ por bartil ¢ 0 gasolco a 20€. Formule o problema de programagao linear que maximize o luo. Capitulo 7 Exereieio 6 da pagina 91 Considere uma emprest que produz dois tipos de fentitizantes, fabricados apenas com silicone ¢ nitrogénio. © fertilizante 1 tem que ter pelo menos 40% de nittogénio © & vendido por T0€Ky. O fenilizante 2 tem que ter pelo menos 70% de silicone e vendido por 406%kg. A empresa pode comprar até 80 kg de nitrogénio a 15€ ‘ky € 100 kg de silicone a 10€/kg. Assumindo que todo o fertilizante produzido pode ser vendido, formalize o problema que maximize os lucros da empresa,Planeamento e Gestio de Projectos Capitulo 8 Exereicio 2 da pagina 180 Use o simplex para resolver o seguinte problema de programagio linear: max 2 = 2x, +x, sh 3x tm $6 Xx, <4 x20, sem restrigdo de sin al Capitulo 9 Capitulo 10 Exereicio 3 da pagina 189 Use 0 método simptex em duas fases para resolver o seguinte problema de programagdo linear: max 2 = Sx,— St 2x +x, =6 x tx, S4 xi +2y, 85 Xk, 20| Exam? 3.1 et Solution Introduction to Linear Programming line pogeraning (iso fr sakng opinion pels, 1947, George Do devel en een met, he sigex lath, slg near oromning ples oko cal LF), Sie te deepen he ile ty, LP bs teen seo ‘ave etinzaton poles in insti os ds stk, efor, fees, pele, en tucking, fa suney of Fre 500 fins, 85% o those soning shat hey dws nur groguing. Aso menue of he ings of ner goon in ‘penton ws, apres 40 of is ok il be dent nen opening ad lk pcan ees. InSecin 3.1, eb our uyof eer rogamming by cing te gered hvac sored by ol nearing robes. e Sates 32 and 3.3, we rn ow save ghia hs iar erauig ples ate oy to wis. Sahin hese imple (wl gv s some use ns fr sohing re cre US. The veil of te cots exis fo a emote acer pogorig de of ee sans. What Is a Linear Programming Problem? In this section, we introduce linear programming and define some important terms that are used to describe linear programming problems, Giapetto’s Woodearving, Ine., manufactures two types of wooden toys: soldiers and trains. A soldier sells for $27 and uses $10 worth of raw materials, Each soldier thatis manufactured increases Giapetto's variable labor and overhead costs by $14. A tain sells for $21 and uses $9 worth of raw materials. Each train built inereases Giapetto’s variable labor and overhead costs by $10. The manufacture of wooden soldiers and trains requires two types bf skilled labor: carpentry and finishing. 4 soldier requires 2 hours of finishing labor andl 1 hour of carpentry labor. A train requires 1 hour of finishing and | hour of carpentry Tabor, Fach week. Giapetto can obiain all the needed raw material but only 100 finishing hours and 80 carpentry hours. Demand for trains is unlimited, but at most 49 soldiers are housht each week. Giapetto wants to maximize weekly profit (revenues ~ costs). Formulate a ‘mathematical model of Giapetto’s situation that can he used to maximize Giapetto’s weekly pratt In developing the Giapetto model, we explore charaeteristies shared by all linear program= ‘ming problems, Decision Variables We bevin by defining the relevant decision variables. In any linear pro. gramming model, the decision variables should completely describe the decisions £0 be50 Chapter 3 nods t Lear Pogerning made (in this ease, by Giapetto). Clearly, Giapetto must decide how many soldiers and ‘rains should be manufactured each week. With this in mind, we define x, = number of soldiers produced each week 1 = number of trains produced each week ObjeciveFundion In any linear programming problem, the decision maker wants to maximize (usually revenue oF profit) or minimize (usually costs) some function of the decision variables. ‘The function to be maximized or minimized is called the objeetive function. For the Giapetto problem, we note that fied costs (such as rent and insurance) do not depend on the values of, andr. Thus, Giapetto can concentrate on maximizing (weekly revenues) — (raw material purchase costs) — (other variable costs) Giapeto’s weekly revenues and costs can be expressed in terms of the decision variables x, and x,. It would be foolish for Giapetto to manufacture more soldiers than ean be sold, so we assume that all toys produced will be sold. Then Weekly evenues = wockly revenues from soldiers + weekly events fom trans = weekly revenues from soldiers = (le) (ea (ssi = (Stee) (a) +S) (Sea =e, + 2x, Also, Weekly raw material costs 0, +9, Other weekly variable costs = I4x, + 10x, ‘Then Giapetto wants to maximize (27x, +2143) — (10x, +985) — Us, + 108) = 35, + Another way to see that Giapetto wants to maximize 3x, + 2x, is to note that Weekly revenues ~ weekly nonfixed cost ‘weekly contribution to profit from soldiers + weekly contribution to profit from trains C= to profit ) (at) sodier week) ceuutine me) (tins) ‘Sa week Also, 10 profit “Soldier Contribution to profit ae Then, as before, we obtain Weekly revenues — weekly nonfixed costs = 3x, +31, Whats alien Prgonring Fem? $Y ‘Thus, Giapetto’s objective is to choose x, and x, to maximize 3x, +2. We use the variable >to denote the objective function value of any LP. Giapeto’s objective function is Maximize = = 3x, +20, o {Un the future, we will abbreviate “maximize” by sae and “minimize” by min.) The coetti- cient of a variable in the objective function is called the objective function coefficient of the vaviable, For example, the objective function coefficient for x, is 3, and the objective function coefficient for x, is 2. In this example (and in many other problems), the objec: tive function coefficient for each variable is simply the contribution of the variable t0 the ‘company’s profit, Constraints Asx, and x, increase, Giapetto’s objective function grows larger. This means that iF Giapetto Were free to choose any values for.x, and.x,. the company could make an arbitrarily large profit by choosing «and x, to be very large. Unfortunately. the values of and x) are limited by the following three restrictions (often called constraints): Constraint 1 Bach week, no more than 100 hours of finishing time may be used, Constraint 2 Each week. no more than 80 hours of carpentry time may be used Constraint 3 Because of limited demand, ut most 40 soldiers should be produced each week. The amount of raw material available is assumed to be unlimited, so 90 restrictions have been placed on this. ‘The next stepin formulating a mathematical model ofthe Giapetto problem is to express Constraints 1-3 in terms of the decision variables 1 and x. To express Constraint 1 in terms of x, and x5, note that Total es stig bs. (seine ) cae mess) soldier ‘week fains ) 2vj tay = 100 a Note that the units of each term in (2) are finishing hours per week, Fre constraint to be reasonable, all terms in the constraint must have the sane units, Otherwise one is adding apples and oranges, ad the constraint won't have a To express Constraint 2 in terms of, and x, note that (az) C2) +(e) (sex) = lx) +1 meaning, Total Phen Constraint 2 may be weitten as vy $y S80 °52 Clapler3—_ emu ie: Rogar Again, note thatthe units of each term in (3) are the same (in this ease, carpentry hours per weeky. Finally, we express the fact that at most 40 soldiers per week can be sold by limiting the ‘weekly production of soldiers to at most 40 soldiers, This yields the following constraint: 1 340 0 Thus (2)-(4) express Constraints 1-3 in terms of the decision variables; they are called ‘constraints for the Giapetto linear programming problem, The coefficients of the deci- sion variables in the constraints are called technological coefficients, This is because the technological coefficients often reflect the technology used to produce different products, For example, the technological coefficient of x, in (3) is 1, indicating that a soldier requires 1 carpentry hour. The number on the right-hand side of each constraint is called the con- saint's right-hand side (or rhs), Often the rhs of a constraint represents the quantity of a resource that is available, SignResicions To complete the formulation ofa linear programming problem, the follow ing question must be answered for each decison variable: Can the decision variable oaly assume nonnevaive values, rs the decision variable allowed to assume both postive and Ifadecison variable x, can only assume nonnegative vlues, wea the ign restretion x, 2 0.1 a variable, ean assume both postive and negative (or zero) values, we say that ¥ is umrestrcted in sig (otten abeviated urs). For he Giapetto problem, itis le that s, 2 O and x, = 0. In ater problems, however. some variables may be urs. For example ifs, represented a firm's eash balance, could be considered negative ifthe ri ote !mote money than iad on hand, tn this cae it would be appropriate to classify a6 urs. Other uses of urs variables are discussed in Section 4.12 jan restrictions x, = O and. > O with the objective funtion (1) and Constants 2)-C4 yields the Following optimization model 3x, 42x, (Objective function) 0 subject to (61) f= 100 (Finishing constraint) @ 4) $5 580 (Carpentry constraint) 8 8) $40 Constraint on demand for soldiers) @ 20 resticton ‘5 0 en restriction) 6) ‘Subject 0" (s4.) means thatthe values of the decision variables x, and x must satisfy all the constraints and all the sign restrictions, Before formally defining a line function and linear inequality i programming problem, we define the concepts of linear tai the vahtes of the devi variables, but we ehoase to comer the epost fom the comsrains The sets fo this wil ecome appre hen we tly thewt mt hut low lon 3.1 Whats oligo Paging Pn? 53 A function f(x). %yo028,) OF 8 for some set of constants C1. €3. ,, isa linear funetion if and only if For example, fx.) = is not a linear function of x, and xy is a linear function of x, and xy, but fLt,.2%)) = For any linear funetion f(4).¥5..+-.¥,) and any number b, the inequalities Fj hgeoeo hy) = Band FO Soon ky) 2 are linear inequalities. + Thus, 2x, +31, <3 and 2x, +x, = 3 are linear inequalities. but x}x, 2 3 is not a Tinear inequality A linear programming problem (LP) is an optimization problem for which we do the following: linear funetion of the decision 1 We attempt to maximize (or minimize Variables. The function that is to be maximized or minimized is ealled the objective fimnction. 2. The values of the decision variables ost satisfy a set of constrains. Each constraint must be a finear equation or finear inequality 3A sign restriction is associated wit restriction specifies either that x, must be nonne: each variable. For any variable x, the ive (x, = 0) or that x, may be unrestricted in sign (urs). Since Giapetto's objective function is linear constraints are fin junction of x, and v,,andall of Giapetto’s mming problem. inequalities, the Giapetto problem is a linea Note that the Giapetto problem is typical of a wide class of finear pro in which a decision maker's 1 is to maximize profit subject to limited resources. The Propotonaty and Adit Assumptions The fact that the objective tunetion for an LP must be a linear fumetion of the decision 1 ‘The contribution of the objective function from each decision variable is proportional bbe. For example, the contribution to the objective funtion tion 10 the o the value of the decision vai from making four soldiers (4 3= $12) is exactly four times the cont objective function from making one sold The contribution to the objective function for any variable is independent of the values df the other decision variables. Forexample, no matter what the value of x,.the manufactere of soldiers will always contribute 3x sto the objective function.54) Goylr3 nod ie gaa ‘Analogously, the fact that cach LP constraint must be a Tinear inequality or linear equation has two implications, 1 The contribution of each variable to the left-hand side of each constraint is proportional to the value of the variable, For example, it takes exactly three times as many finishing hours (2 * 3 = 6 finishing hours) to manufacture three soldiers as it takes to manufacture ‘one soldier (2 finishing hours), 2. The contribution of a variable fo the left-hand side of each constraint is independent of the values of the variable. For example, no matter what the value of x,, the manufacture of xy trains uses finishing hours and x, carpentry hours ‘The first implication given in each lists called the Proportionality Assumption of Linear Programming, Implication 2 ofthe firs list implies thatthe value of the objective function is the sum of the contributions from individual variables, and implication 2 of the second list implies that the left-hand side of each constraint is the sum of the contributions from each variable, For this reason, the second implication in each list is called the Additivity Assumption of Linear Programming. For an LP to be an appropriate representation of a real-life situation, the decision variables must satisfy both the Proportionality and Additivity Assumptions. Two other assumptions must also be satisfied before an LP can appropriately represent a real situation, ‘These are the Divisibility and Certainty Assumptions, The Divsiilty Assumption ‘The Divisibility Assumption requires that each decision variable be allowed to assume Fractional values. For example, inthe Giapetto problem. the Divisibility Assumption implies that it is acceptable to produce 1.5 soldiers or 1.63 trains. Since Giapetto cannot actually produce a fractional number of trains or soldiers, the Divisibility Assumption isnot satisfied inthe Giapetto problem. A linear programming problem in which some or all of the variables rust be nonnegative integers is called an integer programming problem. The solution of integer programming problems is discussed in Chapter 9 in many situations where divisibility is not present, rounding off each variable in the ‘optimal LP solution to an integer may yield a wasonable solution. Suppose the optimal solution to an LP stated that an auto company should manufacture 150,000.4 compact ca luring the curtent year: In this case, you could tell the auto company to manufacture 150,000 ‘of 150,001 compact cars and be fairly confident that this would reasonably approximate an ‘optimal production plan, On the other hand, if the number of missife sites that the United States should use were a variable in an LP and the optimal LP solution said that 0.4 missile sites should be built, ic would make a big difference whether we rounded the number of missile sites down to 0 or up t0 1. In this situation, the integer programming methods of Chapter 9 would have to be used, because the number of missile sites is definitely not divisible, The Certainty Assumption The Certainty Assumption is that each parameter (objective funtion coetficient, right- hand side, and technol efficient) is known with certainty, IF we were unsure of the exact amount of earpentry and finishing hours required fo build a train, the Certainty Assumption would be violatedere TA 31 huts oti Payurning abn? 55 Feasible Region and Optimal Solution ‘Two of the most basic concepts associated with a linear programming problem are feasible region and optimal solution, For defining these concepis, we use the term point to mean a specilication of the value for each decision variable. DEFINITION © The feasible region for an LP is the set of all points satistying all the LP's constraints and all the LP'S sign restrictions, = For exan region, Note that x, (5)-(6): le. in the Giapetio problem, the point (x, = 40, x) = 20) is in the feasible }Oand x, = 20satisfy the constraints (2}-(4) and the sign restrictions Constraint (2), 2x, +.t, < 100, is satisfied, because 2440) + 20 = 100, Constraint (3), x, +, < 80, is satisfied, because 40 + 20 = 80. Constraint (,.x, = 40, i satisfied, because 40 < 40, Restriction (5), x, > 0, is satisfied, because 40 > 0, Restriction (6), x, = 0, is satisfied, because 20 > 0, On the other hand, the point (x, = 15, 1, = 70) is notin the feasible region, because even ‘hough x, = 15 and x, = 70 satisty (2), (4), (5), and (6), they fail to satisfy (3}¢ 15 + 70 is hot less than or equal to 80. Any point that is notin an LP's Feasible region is said to be an infeasible point, As anotherexample of an infeasible point, consider (x, = 40, x, = —20). | Although this point satisfies all the constraints and the sign restriction (5), its infeasible because it fails to satisfy the sign restriction (6). > 0. The feasible region for the Giapetto problem isthe set of possible production plans that Giapetto must consider in searching for the optimal production phan. DEFLUL TION © Bora maximization problem, an optimal solution to an LP is a point inthe feasible on with the largest objective function value, Similarly, for a minimization problem, an optimal solution is a point in the feasible region with the smallest objective function value,» Most LPs have only one optimal solution. However, some LPs have no optimal solution, and some LPs have an infinite number of solutions (these situations are discussed in Section 3.3). tm Section 3.2, we show that the unique optimal solution to the Giapetto problem is (x, = 20,.x = 60). This solution yields an objective function value of 2 = Buy +2 = 3120) + 2160) = $180 When we say that (1, = 20,.r, = 60) is the optimal solution to the Giapetto problem, we are saying that no point inthe feasible region has an objective function valuc exceeding 180, Giapetto can maximize profit by building 20 soldiers and 60 trains each week. If Giapetto were to prodice 20 soldiers and 60 trains each week, the weekly profit would be $180 less weekly fixed casts, For example. if Giapetto’s only fixed cost were rent of $100 per week. ly profit would he 180 — 100 = $80 per week56 Cher Problems Group A 1 Farmer Jones must determine how many aeres of com and wheat to plant this year. An acre of wheat yies 25 bushels of wheat and requires 10 hours of labor per week, [An acre of comm yields 10 bushels of earn and requies 4 hours ‘of labor per week. AI! wheat can be soldat Sa bushel, and all cor can be sod at $3 a bushel, Seven aeres of land and +40 hours per week of labor are available. Government regu lations roquite that atleast 30 bushels of corn be produced during the current year. Let x, = number of acres of com Planted, and x, — number of ders of wheat planted. Using ‘hese decision Variables formulae an LP whose solution wil, tell Farmer Jones how to. maximize the total revenve from wheat snd corn, 2 Amvwer these questions about Problem a Is6x, =2, x, = 3) inthe feasible region? boku =key © bey =2e 3) inthe feasible region? =1) inthe feasible region? lord tier Paging 4 Is ts, = 3.4, = 2) nthe feasible gion? 3. Using the variables 1 = number of bushels of corn produced and x, = number of bushels of wheat produced, reformlate Farner Jones's LP, 4 Truckcomanvfactures two types of tucks: [and 2. Each ‘rack must go through the painting shop and assembly shop. IF the painting shop were completely devoted to painting type | trucks, 800 per day could be punted, wherets ifthe printing shop were completely devoted to painting type 2 trucks, 700 per day could be painted. Ifthe assembly shop were completely devoted t assembling truck | engines, 1500 per day could be assembled, and ifthe assembly shop were ‘completely devoted to assembling truck ? engines, 1200 per day could be assembled. Each type I truck conteitutes $300 to profit; each type 2 ruck contributes $500. Formulate an LLP that will maximize Truckeo’s prot Group B 5) Why don’t we allow an LP to have « oF > constraints? 3.2. The Graphical Solution of Two-Variable Linear Programming Problems Any LP with only (wo variables can be salvet phically. We always label the variables Fad NT eRe hes STR Cn Fa Ta 2x, £3x <6 a The same set of points (x,,.x,) satisfies 3x, $6~24, ‘This last inequality may be rewritten as Since moving downward on the graph decreases x that satislies (8) and (7) lies on or below the line wv indicated by darker shading in and 2x, + 3 (on or below the line 2x, ++ 3x, lies on or above the line 2x, Figure 1.) gure 1. Note, however, that x, 6 are all the same Tine, This means that the set of points Satisiying (7) lies 2s o (sce Figure 1), the set of poins 2— fay. This set of pois is By dry =6— 26 Similarly, the set of points satisfying 2x, + 3x, = 6 6, (These points are shown by lighter shading in ‘Consider a linear inequality constraint of the form f(x,..x5) = bor f(y.) , defining the inequality plus all There is an easy way to determine the side of the line for which an inequality such as Fey.) SB oF fl, 4) ~ b is satisfied. Just choose any point P that does not satisfyee FieuRET Gaping ine Iga 2 The Gopi! Soin foYoiablLinewrPoyarng Rebs 57 (RM Satisties 20, + 34, = 6 4 Satisties 26, + 31 = 6 the Hine /'(x,..%) = b. Determine whether P satisfies the inequality. If it does. then all points on the same side as P of fx,,.t5) = 6 will satisfy the inequality. IF P does not satisfy the inequality, then all points on the other side of f'(r,..%) = 6, which does not contain P. will satisfy the inequality. For example, to determine whether 2x, + 3x, > 6 is satisfied by points above or below the line 2x, ++ 3r,=6. we note that (0,0) does not satisfy 2x, ++ 38, = 6, Since (0,0) is below the line 2x, + 3x, = 6, the set of points ¥) +34, 2 6 includes the line 2x, + 3x, =6 and the points above the Hine ‘This agrees with Figure 1 . Finding the Feasible Solution We illustrate how (o solve two-variable LPs graphically by solving the Giapetto problem. To begin, we graphically determine the feasible region for Giapetto's problem, The feasible region for the Giapetto problem is the set of all points (x, x) satisfying 2x, $4, S100 (Constraims) a ay try £80 @ ® (Sign restrictions) 6 6) For a point (x,..r,) to be in the feasible region, (x,,.x5) must satisfy alf the inequalities (2)-6), Note that the only points satisfying (5) ane! (6) lie inthe first quadrant of the y= plane. This is indicated in Figure 2 by the arrows pointing to the right fromm the 1, axis and upward from the x, axis. Thus, any poin that is outside the first quadrant cannot be in |FIBURE 2 GephilSlon of Gopeto Pale 58° Chaves tnd to oar Fogonning ‘M288 Feasible region the feasible region. This means thatthe feasible region will be the set of points in the frst ‘quadrant that satisfies (2)-(4). (Our method for determining the set of points that satisfies a linear inequality will alse identify those that meet (2)-(4). From Figure 2, we see that (2) is satisfied by all points below or on the line AB (AB is the line 2x, +x, = 100). Inequality (3) is satisfied by all points on or below the line CD (CD is the line x, +x, = 80). Finally, (4) is satisfied by all points on orto the left of line EF (EF is the line x, = 40). The side ofa fine that satisfies an inequality is indicated by the direction of the arrows in Figure 2. From Figure 2, we see that the set of points in the first quadrant that satisfies (2), (3), and (4) is bounded by the five-sided polygon DGFEH. Any point on this polygon or in its interior isin the feasible region. Any other point fails to satisfy at least one of the inequalities (2)-(6). For example, the point (40, 30) lies outside DG FEH because it is above the line segment A B. Thus (40, 30) is infeasible, because it fuls to satisfy (2). An easy way 10 find the feasible region is 10 determine the set of infeasible points, Note that all points above line AB in Figure 2 are infeasible, because they fail to satisfy ). Similarly, all points above CD are infessible, because they fail 10 satis'y (3), Also, all points to the Fight of the vertical line £F are infeasible, because they fail to satisfy (4, Afier these points are eliminated from consideration, we are left with the feasible region (DGFEH) Finding the Optimal Solution Having identified the feasible region for the Giapetto problem, we now search for the ‘optimal solution, which will be the point in the feasible region with the largest value ofDEFINITION erin Tow {LD The Gopi Salt of Tuoi ins Prgrning Palos 59 ‘To find the optimal solution. we need to graph a line on which all points flue. In a max problem, such a line is called an isoprofit Hine (in & have the same min problem, an isocost line.) To draw an isoprofit line, choose any point in the feasible region and calculate its ¢-value, Let us choose (20, 0). For (20. 0), : = 320) + 2(0) = 60. Thus, (20,0) lies on the isoproit line = = 3x, +2 = 60. Rewriting 34, + 25 = 60 as vy, = 30 — $x, We see that the isoprofit line 3x, + 2x, = 60 has a slope of ~. Since all isoprit ines re of the form 3x, + 2x, = constant, al isoprofit Fines have the same slope. ‘This means thar once we have drawn one isoprofit line, we can fad all other isopropi lines by moving parallel tothe isoprofit line we have drawn. Itis now clear how to find the optimal solution toatwo-variable LP. After you have drawn isoprofit line, generate other isoprofit Hines by moving parallel tothe drawn isoprofit line in a direetion that inereases : (for a max problem). After a point the isoprofit lines will no longer intersect the feasible region. The last isoprofit line intersecting (touching) the feasible region defines the largest z-value of any point in the Feasible region and indicates the optimal solution to the LP. In our problem, the objective function z = 3x, + 2x, will inerease if we move in a direction for which both x, and x, inetease. Thus, we construct additional isoprofit lines by moving parallel to 3x, + 2x, = 60 in a northeast direction (upward and (o the right), From Figure 2, we see that the isoprofit line passing through point G is the last isoprotit line to interscet the feasible region. Thus, G is the point in the feasible region with the largest z-value and is therefore the optimal solution to the Giapetto problem. Note that point G is where the lines 2x, +x, = 100 and x, +x, = 80 imersect, Solving these two equations simultaneously. we find that (x, = 20. the optimal solution to the Giapetto problem. The optimal value of z may be found by substituting these values of x, ands, into the objective funetion. Thus, the optimal value of is ¢ = 3(20) + 2060) = 180. Binding and Nonbinding Constraints Once the optimal solution to an LP has been found, itis useful (see Chapters 5 and 6) to classify each constraint as being a binding constraint or a nonbinding constraint. A constraint is binding i the left-hand side and the right- hand side of the constraint fre equal when the optimal values of the decision variables are substituted into the const ‘Thus, (2) and (3) are binding constraints A constraint is nonbinding if the left-hand side snd the right-hand side of the constraint are unequal when the optimal values of the decision variables are substituted into the constraint, * Since x, = 20 is less than 40, (4) is a nonbind sonstrant Convex Sets, Extreme Points, and LP The feasible region for the Giapetio problem is an example of a convex setDEFINITION DEFINITION 60 Gaptr loko Une oganeg A set of points S is a convex set if the line segment joining any pair of points in is wholly contained in S. Figure 3 gives four illustrations of this definition. In Figures 3a and 3b, each line segment joining two points in S contains only points in S. Thus, in both these figures, ‘convex. In Figures 3c and 3d, 5 is not convex. In each figure, points A and B are in S, but there are points on the Tine segment A that are not contained in S. fn our study of linear programming, a certain type of point in a convex set (called an extreme point) is of great interest. For any convex set S, a point P in S is an extreme point if each line segment that lies completely in $ and contains the point P has P as an endpoint of the line segment. * For example. in Figure 3a, each point on the circumference of the circle is an extreme point of the circle. In Figure 3b, points A, B.C, and D are extreme points of S. Although point £ is on the boundary of S in Figure 3b, E is not an extreme point of S. This is because E lies on the line segment AB (AB lies completely in $), and £ is not an endpoint of the line segment 4 B., Extreme points are sometimes called earner points, because if the set isa polygon, the extreme points of S will be the vertices, or corners, of the polygon, The feasible region for the Giapetto problem is a convex set. This is no accident: It can be shown that the feasible region for any LP will be a convex set, From Figure 2, we see ‘that the extreme points ofthe feasible region are simply points D, F, E, Gand H. Ttean be shown thatthe feasible region for any LP has only a finite number of extreme points, Also ‘note that the optimal solution to the Giapetto problem (point G) is an extreme point of the feasible region. It can be shown that any LP that has an optimal solution has an extreme point that és optimal. This result is very important, because it reduces the set of points ‘yielding an optimal solution from the entire feasible region (which generally contains 3 infinite number of points) to the set of extreme points (a finite set). For the Giapetto problem, its easy to see why the optimal solution must be an extreme point ofthe feasible region. We note that z increases as we move isoprofit lines in a northeast direction, so the largest z-value in the feasible region must occur at some point P that bas no points in the Feasible region northeast of P. This means that the optimal solution must FAG URE 3 Comer od Nowe Sis 5 = shaded area 4 BEXAMPLE 2 Solaton LD The Bopha Soluce of aoe ier Progamming Pothms 61 lie somewhere on the boundary of the feasible region DGFEH. The LP must have an ‘extreme point that is optimal, because for any line Segment on the boundary of the Feasible region, the largest z-value on that line segment must be assumed at one of the endpoints of the line segment ‘To sce this, look atthe line segment FG in Figure 2. FG is pattof the line 2x, ++ x, 100 and has slope of ~2. If we move along FG and decrease x, by I, then s by 2. and the value of z changes as follows: 3x, goes down by 3(1) = 3and 2x, goes up by 2(2) = 4. Thus, in total, z increases by 4 — ‘This means that moving along FG in a direction of decreasing x, increases =. Thus. the valuc of 2 at point G must exceed the value of = at any other point on the line segment FG. A similar argument shows that for any objective function, the maximum value of : on 1 given line segment must occur at an endpoint of the line segment. Therefore, for any LP, the largest z-value in the feasible region must be attained at an endpoint of one of the line segments forming the boundary of the feasible region. In short, one of the extreme points, of the feasible region must be optimal. (To test understanding, the reader should show that if Giapetto’s objective function were : = 6x, +5, point F would be optimal, whereas it Giapetto’s objective function were z = x, + 64,, point D would be optimal.) ‘Our proof that an LP alway’ has an optimal extreme point depended heavily on the fact that both the objective function and the constraints were linear functions, tn Chapter 12, wwe show that for an optimization problem in which the abjective function or some of the constraints are not linear. the optimal solution to the optimization problem may not occur a an extreme point The Graphical Solution of Minimization Problems Dorian Auto manufactures luxury cars and trucks, The company believes that its most likely customers are high-income women and men. To reach these groups, Dorian Auto hats embarked on an ambitious TV advertising campaign and has decided to purchase I-minute commercial spots on two types of programs: comedy shows and football games, comedy commercial is seen by 7 million high-income women and 2 million high-income men, Each football commercial is seen by 2 million high-income women and 12 million high-income men. A [-minute comedy ad costs $50,000, and a I-minute football ad costs '$100,000, Dorian would like the commercials to be seen by at least 28 million high-income women and 24 million high-income men, Use linear programming to determine how Dorian Auto can meet its advertising requirements at minimum cost ch Dorian mustdeeide hoon variables are ny comedy and football ads should be purchased. sothe decision x, = number of |-minute comedy ads purchased ‘ry = number of L-minute football ads purchased Then Dorian wants to minimize to 1 advertising cost (in thousands of dollars). Toual advertising cost = cost of comedy ads. + cost af football ads = (ea) (eomey as) + (rasan) (rasta as) = 50s, + 100K,