0 ratings0% found this document useful (0 votes) 2K views12 pagesGoal Programming
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content,
claim it here.
Available Formats
Download as PDF or read online on Scribd
8.1
CHAPTER 8
Goal Programming
The LP models presented in the preceding chapters are based on the optimization of a
single objective function. There are situations where multiple (possibly conflicting)
objectives may be more appropriate. For example, aspiring politicians may promise to
reduce the national debt and, simultaneously, offer income tax relief. In such situations,
it may be impossible to find a single solution that optimizes the conflicting objectives.
Instead, we may seek a compromise solution based on the relative importance of each
objective.
This chapter presents the goal programming technique for solving multiobjective
models. The principal idea is to convert the original multiple objectives into a single
goal. The resulting model yields what is usually referred to as an efficient solution
because it may not be optimum with respect to all the conflicting objectives of the
problem.
A GOAL PROGRAMMING FORMULATION
The idea of goal programming is illustrated by an example.
Example 8.1-1
Fairville is a small city with a population of about 20,000 residents. The city council is in
the process of developing an equitable city tax rate table. The annual taxation base for
real estate property is $550 million. The annual taxation bases for food and drugs and
for general sales are $35 million and $55 million, respectively. Annual local gasoline
consumption is estimated at 7.5 million gallons. The city council wants to develop the
tax rates based on four main goals.
1. Tax revenues must be at least $16 million to meet the city’s financial commit-
ments.
2. Food and drug taxes cannot exceed 10% of all taxes collected. «
347348 Chapter8 Goal Programming
3. General sales taxes cannot exceed 20% of all t:
4, Gasoline tax cannot exceed 2 cents per gallon.
collected.
Let the variables x,, x,,and x, represent the tax rates (expressed as proportions of
taxation bases) for property, food and drugs, and general sales; and define the variable
x, as the gasoline tax in cents per gallon. The goals of the city council are then
expressed as
5950x, + 35x; + 55x, + O75x, = 16 (Tax revenue)
35x, S .1(550x, + 35x; + 55x, + .075x,) _ (Food/drug tax)
55x, = .2550x, + 35x, + 55x, + .075x,) (General tax)
X, <2 (Gasoline tax)
Xp Xp Xu Xp ZO
‘These constraints are then simplified as
50x, + 35x; + 55x, + .075x, = 16
55x, — 31.5x, + 5.5x, + .0075x, = 0
0x, + 7x) — 44x, + OL
ss
Xp Xp Xo Xp =O
Each of the inequalities of the model represents a goal that the city council aspires
to satisfy. Most likely, however, the best we can do is seek a compromise solution
among these conflicting goals.
The manner in which goal programming finds a compromise solution is to convert
each inequality into a flexible goal in which the corresponding constraint may be vio-
lated, if necessary. In terms of the Fairville model, the flexible goals are expressed as
follows:
550x, + 35x, + 55x, + .075x, + sf - sy = 16
55x, — 31.5x, + 5.5x, + .0075x, + 53 —
10x, + Tx, — 44x, + O15, + 5} — 53 = 0
x, +s} - 55 = 2
Xp. Xp Xo Xy = 0
st, sj 2 0,i = 1,2,3,4
The nonnegative variables s/ and s;, i = 1, 2, 3, 4, are called deviational variables
because they represent the deviations above and below the right-hand side of con-
straint i.
The deviational variables s* and s; are by definition dependent and, hence, cannot
be basic variables simultaneously. This means that in any simplex iteration, at most one
of the two deviational variables can assume a positive value. If the original ith inequal-
ity is of the type = and its s7 > 0, then the ith goal will be satisfied; otherwise, if s; > 0.
goal i will not be satisfied. In essence, the definition of s; and s; allows us to meet or vio-8.1 AGoal Programming Formulation 349
late the ith goal at will. This is the type of flexibility that characterizes goal program-
ming when it seeks a compromise solution. Naturally, a good compromise solution aims
at minimizing the amount by which each goal is violated.
In the Fairville model, given that the first three constraints are of the type =
and the fourth constraint is of the type =, the deviational variables sj, 53, s3, and sz of
the problem represent the amounts by which the respective goals are violated. Thus,
the compromise solution tries to satisfy the following four objectives as much as
possible:
Minimize G, = sj
Minimize G, = s$
Minimize G, = s;
Minimize G, = si
‘These functions are minimized subject to the constraint equations of the model.
How can we optimize a multiobjective model with possibly conflicting goals?
Two methods have been developed for this purpose: (1) the weights method and
(2) the preemptive method. Both methods are based on converting the multiple objec-
tives into a single function as detailed in Section 8.2.
PROBLEM SET 8.14
1. Formulate the Fairville tax problem, assuming that the town council is specifying an addi-
tional goal, Gs, that requires gasoline tax to equal at least 10% of the total tax bill.
2. The NW Shopping Mall conducts special events to attract potential patrons. The two most
popular events that seem to attract teenagers, the young/middle-aged group, and senior
citizens are band concerts and art and craft shows. The costs per presentation of the band
and art show are $1500 and $3000, respectively. The total (strict) annual budget allocated
to the two events is $15,000. The mall manager estimates the attendance of the events as
follows:
Number attending per presentation
Event Teenagers Young/middle age __ Seniors
Band concert 200 100 0
Art show 0 400 250
‘The manager has set the minimum annual goals of 1000, 1200, and 800 for the attendance
of teenagers, the young/middle-aged group, and seniors, respectively. Formulate the prob-
lem as a goal programming model.
3. Ozark University admissions office is processing freshman applications for the upcoming
academic year. The applications fall into three categories: instate, out-of-state, and inter-
national. The male-female ratios for in-state and out-of-state applicants are 1:1 and
3:2, respectively. For the international students, the corresponding ratio is 8:1. The
‘American College Test (ACT) score is an important factor in accepting new students350
Chapter 8 Goal Programming
4
Statistics indicate that the average ACT scores for in-state, out-of-state, and international
students are 27, 26, and 23, respectively. The committee on admissions has established the
following desirable goals for the new freshman class
(a) The incoming class is at least 1200 freshmen.
(b) The average ACT score for all incoming students is at least 25.
()_ International students constitute at least 10% of the incoming class.
(a) The female-male ratio is at least 3:4
(e) Out-of-state students constitute at least 20% of the incoming class.
Formulate the problem as a goal programming model
Circle K farms consume 3 tons of special feed daily. The feed—a mixture of limestone,
corn, and soybean meal—must satisfy the following nutritional requirements:
Calcium. Atleast 0.8% but not more than 1.2%
Protein. Atleast 22%
Fiber. At most 5%
The following table gives the nutritional content of the feed ingredients.
Ib per Ib of ingredient
Ingredient Calcium — Protein Fiber
Limestone 380 00 00
Corn 001 09 02
Soybean meal 002 50 08
Formulate the problem as a goal programming model, and state your opinion regarding
the applicability of goal programming to this situation.
Mantel produces a toy carriage, whose final assembly must include four wheels and two
seats, The factory producing the parts operates three shifts a day. The following table pro-
vides the amounts produced of each part in the three shifts.
Units produced per run
Shift Wheels Seats
1 500 300
2 600 280
3 640 360
Ideally, the number of produced wheels is exactly twice that of the number of seats.
However, because the production rates vary from shift to shift, exact balance in pro-
duction may not be possible. Mantel is interested in determining the number of pro-
duction runs in each shift that minimizes the imbalance in the production of the parts.
‘The capacity limitations restrict the number of runs to between 4 and 5 for shift 1, 10
and 20 for shift 2, and 3 and 5 for shift 3. Formulate the problem as a goal program.
ming model.
Camyo Manufacturing produces four parts that require the use of a lathe and a drill
press. The two machines operate 10 hours a day. The following table provides the time in
minutes required by each part:8.1 AGoal Programming Formulation 351
Production time in min
Part Lathe Drill press
1 s 3
2 6 2
3 4 6
4 7 4
It is desired to balance the two machines by limiting the difference between their total
operation times to at most 30 minutes. The market demand for each part is at least 10
units. Additionally, the number of units of part 1 may not exceed that of part 2. Formulate
the problem as a goal programming model.
‘Two products are manufactured on two sequential machines, The following table gives
the machining times in minutes per unit for the two products.
Machining time in min
Machine Product! Product 2
1 5
6
The daily production quotas for the two products are 80 and 60 units, respectively. Each
machine runs 8 hours a day. Overtime, though not desirable, may be used if necessary to
meet the production quota. Formulate the problem as a goal programming model.
8. Vista City Hospital plans the short-stay assignment of surplus beds (those that are not
already occupied) 4 days in advance. During the 4-day planning period about 30,25, and
20 patients will require 1-,2-, or 3-day stays, respectively. Surplus beds during the same
period are estimated at 20,30, 30, and 30. Use goal programming to resolve the problem
of overadmission and underadmission in the hospital.
9. The Von Trapp family is in the process of moving to a new city where both parents have
accepted new jobs. In trying to find an ideal location for their new home, the Von Trapps
list the following goals:
(a) It should be as close as possible to Mrs. Von Trapp’s place of work (within } of a
mile).
(b) It should be as far as possible from the noise of the airport (at least 10 miles).
(©) It should be reasonably close to a shopping mall (within 1 mile). co
Mr. and Mrs. Von Trapp use a landmark in the city as a reference point and locate the
x-y coordinates of work, airport, and shopping mall at (1,1), (20, 15), and (4,7), respec-
tively (all distances are in miles). Formulate the problem as a goal programming model
(Note: The resulting constraints are not necessarily linear.)
10. Regression Analysis. In a laboratory experiment, suppose that y; is the ith observed
(independent) yield associated with the dependent observational measurements
Xi = 1,2, ..., m;j = 1,2, ..., n-Itis desired to determine a linear regression fit into
these data points. Given b,, j = 0, 1, ..., ,as the regression coefficients, all b, are deter-
mined such that the sum of the absolute deviations between the observed and the esti-
mated yield is minimized. Formulate the problem as a goal programming model.352
8.2
8.2.1
Chapter 8 Goal Programming
11. Chebyshev Problem. An alternative goal for the regression model in Problem 10 is to
minimize over b, the maximum of the absolute deviations. Formulate the problem as a
goal programming model
GOAL PROGRAMMING ALGORITHMS
This section presents two algorithms for solving goal programming. Both methods con-
vert the multiple goals into a single objective function. In the weights method, the sin-
gle objective function is the weighted sum of the functions representing the goals of
the problem. The preemptive method starts by prioritizing the goals in order of impor-
tance, The model is then optimized using one goal at a time such that the optimum
value of a higher priority goal is never degraded by a lower priority goal.
‘The proposed two methods do not generally produce the same solution. Neither
method, however, is superior to the other because each technique is designed to satisfy
certain decision-making preferences.
The Weights Method
Suppose that the goal programming model has n goals and that the ith goal is given as
Minimize G,, i = 1, 2, ...,."
The combined objective function used in the weights method is defined as
Minimize z = w,G, + w,G, + + w,G,,
The parameter w,, i = 1, 2, ..., n, represents positive weights that reflect the decision
maker’s preferences regarding the relative importance of each goal. For example,
w, = 1, for alll i, signifies that all goals carry equal weights. The determination of the
specific values of these weights is subjective. Indeed, the apparently sophisticated ana-
lytic procedures developed in the literature (see, e.g., Cohon, 1978) are still rooted in
subjective assessments.
Example 8.2-1
TopAd, a new advertising agency with 10 employees, has received a contract to pro-
mote a new product. The agency can advertise by radio and television. The following
table provides data about the number of people reached by each type of advertise-
ment, and the cost and labor requirements.
Data/min advertisement
Radio Television
Exposure (in millions of persons) 4 8
Cost (in thousands of dollars) 8 4
Assigned employees 1 2
The contract prohibits TopAd from using more than 6 minutes of radio advertisement.
Additionally, radio and television advertisements need to reach at least 45 million peo-8.2 Goal Programming Algorithms 353
ple. TopAd has set a budget goal of $100,000 for the project. How many minutes of
radio and television advertisement should TopAd use?
Let x, and x, be the minutes allocated to radio and television advertisements. The
goal programming formulation for the problem is given as
Minimize G, = s} (Satisfy exposure goal)
Minimize G, = s; (Satisfy budget goal)
subject to
4x, + 8x2 + sf — 55 = 45 (Exposure goal)
8x, + 24x, + s$ — sy = 100 (Budget goal)
xt 2x = 10 (Personnel limit) .
x, = 6 (Radio limit)
20
TopAd’s management assumes that the exposure goal is twice as important as the
budget goal. The combined objective function thus becomes
Minimize z = 2G, + G, =
The optimum solution (obtained by TORA) is
z= 10
x, = 5 minutes, x, = 2.5 minutes, s} = 5 million persons
Xp Xp St, 57
All the remaining variables equal zero.
The fact that the optimum value of z is not zero indicates that at least one of the
goals is not met. Specifically,s; = 5 means that the exposure goal (of at least 45 million
persons) is missed by 5 million individuals. Conversely, the budget goal (of not exceed-
ing $100,000) is not violated because s; = 0.
Goal programming yields only an efficient solution to the problem, which is not
necessarily optimum. For example, the solution x, = 6 and x; = 2 yields the same
exposure (4 x 6 + 8 X 2 = 40 million persons) but costs less (8 X 6 + 24 x 2=
$96,000). In essence, what goal programming does is to find a solution that simply
satisfies the goals of the model with no regard to optimization. Such “deficiency” in
finding an optimum solution raises doubts about the viability of goal programming as.
an optimizing technique (see Example 8.2-3 for further discussion).
PROBLEM SET 8.2A
1. Consider Problem 1, Set 8.1a dealing with the Fairville tax situation. Solve the problem,
assuming that all five goals have the same weight. Does the solution satisfy all the
goals?
2. In Problem 2, Set 8.1a, suppose that the goal of attracting young/middle-aged people is
twice as important as either of the other two categories (teens and seniors). Find the asso-
ciated solution, and check if all the goals have been met.
3. In the Ozark University admission situation described in Problem 3, Set 8.1a, suppose
that the limit on the size of the incoming freshman class must be met, but the remaining354
8.2.2
Chapter 8 Goal Programming
requirements can be treated as flexible goals. Further, assume that the ACT score goal is
twice as important as any of the remaining goals.
(a) Solve the problem, and specify whether or not all the goals are satisfied.
(b) If,in addition, the size of the incoming class can be treated as a flexible goal that is
twice as important as the ACT goal, how would this change affect the solution?
In the Circle K model of Problem 4, Set 8.1a, is it possible to satisfy all the nutritional
requirements?
5, In Problem 5, Set 8.1a, determine the solution, and specify whether or not the daily pro-
duction of wheels and seats can be balanced.
In Problem 6, Set 8.1a, suppose that the market demand goal is twice as important as that
of balancing the tw6 machines, and that no overtime is allowed. Solve the problem, and
determine if the goals are met.
7. In Problem 7, Set 8.1a, suppose that the production quota for the two products needs to
be met, using overtime if necessary. Find a solution to the problem, and specify the
amount of overtime, if any, needed to meet the production quota.
In the Vista City Hospital of Problem 8, Set 8.1a, suppose that only the bed limits repre-
sent flexible goals and that all the goals have equal weights. Can all the goals be met?
‘The Malco Company has compiled the following table from the files of five of its employ-
ces to study the relationship between income and age, education (expressed in number of
college years completed), and experience (expressed in number of years in the business)
4
8
%
‘Age (yr) Education (yr) Experience (yr) Annual income ($)
30 4 5 40,000
39 5 10 48,000
44 2 4 38,000
48 0 18 36,000
37 3 9 41,000
Use the goal programming formulation in Problem 10, Set 8.1a to fit the data into
the linear equation y = by +.b,x; + byxy + byX
10. Solve Problem 9 using the Chebyshev Method proposed in Problem 11, Set 8.1a
The Preemptive Method
In the preemptive method, the decision maker must rank the goals of the problem in
order of importance. Given an n-goal situation, the objectives of the problem are writ-
ten as
Minimize G, = p, (Highest priority)
Minimize G,, = p, (Lowest priority)
The variable p; is either s* or s; representing goal i, For example, in the TopAd model
(Example 8.2-1),p; = sj and p; = 53.
The solution procedure considers one goal at a time, starting with the highest
priority, G,, and terminating with the lowest, G,. The process is carried out such that8.2 Goal Programming Algorithms 355
the solution obtained from a lower priority goal never degrades any higher priority
solutions.
The literature on goal programming presents a “special” simplex method that
guarantees the nondegradation of higher priority solutions. The method uses the
column-dropping rule that calls for eliminating a nonbasic variable x, with z, — c, # 0
from the optimal tableau of goal G, before solving the problem of goal G,.,;.The rule
recognizes that such nonbasic variables, if elevated above zero level in the optimiza-
tion of succeeding goals, can degrade (but never improve) the quality of a higher prior-
ity goal. The procedure requires modifying the simplex tableau so that it will carry the
objective functions of all the goals of the model.
The proposed column-dropping modification needlessly complicates goal pro-
gramming. In this presentation, we show that the same results can be achieved in a
more straightforward manner using the following steps:
Step 0. Identify the goals of the model and rank them in order of priority:
G, =p > G.=—) > > Gn = Pp
Seti=1
Solve LP, that minimizes G,, and let p, = p; define the corresponding opti-
mum value of the deviational variable p,. If i = n, stop; LP, solves the n-goal
program. Otherwise, augment the constraint p, = p; to the constraints of the
G;-problem to ensure that the value of p, will not be degraded in future prob-
lems. Set i = i + 1, and repeat step i.
Step
The successive addition of the special constraints p, = p; may not be as “elegant”
theoretically as the column-dropping rule. Nevertheless, it achieves the exact same
result. More important, it is easier to understand.
Some may argue that the column-dropping rule offers computational advan-
tages. Essentially, the rule makes the problem smaller successively by removing vari-
ables, whereas our procedure makes the problem larger by adding new constraints.
However, considering the nature of the additional constraints (p, = p;), we should be
able to modify the simplex algorithm to implement the additional constraint implicitly
through direct substitution of the variable p;. This substitution affects only the con-
straint in which p, appears and, in effect, reduces the number of variables as we move
from one goal to the next. Alternatiyely, we can use the bounded simplex method of
Section 7.3 by replacing p; = p; with p, = p;, in which case the additional constraints
are accounted for implicitly. In this regard, the column-dropping rule, theoretical
appeal aside, does not appear to offer a particular computational advantage. For the
sake of completeness, however, we will demonstrate in Example 8.2-3 how the column-
dropping rule works.
Example 8.2-2
The problem of Example 8.2-1 is solved by the preemptive method. Assume that the
exposure goal has a higher priority.356 Chapter 8
Step 0.
Step 1.
Step 2.
Goal Programming
G, > G;
G): Minimize s} (Satisfy exposure goal)
Gy: Minimize s; (Satisfy budget goal)
Solve LP).
Minimize G, = sf
subject to
4x, + 8x, (Exposure goal)
8x, + 24x, (Budget goal)
x + 2x (Personnel limit)
x 6 (Radio limit)
1X, XS}, Sh Sh, 8] = 0
‘The optimum solution (determined by TORA) is x, = 5 minutes, x» = 2.5
minutes,s} = 5 million people, with the remaining variables equal to zero. The
solution shows that the exposure goal, G,, is violated by 5 million persons.
In LP), we have p; = s}. Thus, the additional constraint we use with the
G,-problem is s}
We need to solve LP, whose objective function is
Minimize G, = s3
subject to the same set of constraints as in step 1 plus the additional con-
straint st = 5. We can solve the new problem by using TORA’s MODIFY
option to add the constraint s} = 5.
The additional constraint sy = 5 can also be accounted for by substituting
out sf in the first constraint. The result is that the right-hand side of the expo-
sure goal constraint will be changed from 45 to 40, thus reducing LP; to
Minimize G, = 53
subject to
4x, + 8x - 5) 40 (Exposure goal)
8x, + 24x, + si — sj = 100 (Budget goal)
x, + 2x) = 10 (Personnel limit)
x = 6 (Radio limit)
Xiy Xa) ST 83, 57 = 0
The new formulation is one variable less than the one in LP,, which is the
general idea advanced by the column-dropping rule.
In reality, the optimization of LP, is not necessary in this example because
the optimum solution to problem G, already yields s; = 0. Hence, the solu-
tion of LP, is automatically optimum for LP, as well (you can verify this
answer by solving LP, with TORA).8.2 Goal Programming Algorithms 357
Next, we use an example to show that a better solution for the problem of
Example 8.2-2 can be obtained if the preemptive method is used to optimize objectives
rather than to satisfy goals. The example also serves to demonstrate the column-
dropping rule for solving goal programs.
Example 8.2-3
The goals of Example 8.2-2 can be restated as.
Priority 1: Maximize exposure (P,)
Pwiority 2: Minimize cost (P,)
Mathematically, the two objectives are given as
Maximize P, = 4x, + 8x, (Exposure)
Minimize P, = 8x, + 24x, (Cost)
The specific goal limits for exposure and cost ( = 45 and 100) are removed because the
simplex method will determine them optimally.
The new problem can thus be stated as
Maximize P, = 4x, + 8x,
Minimize P, = 8x, + 24x,
subject to
x, + 2x, = 10
x =6
X10 = 0
We first solve the problem using the procedure introduced in Example 8.2-2
Step 1. Solve LP).
Maximize P, = 4x, + 8x)
subject to
x, + 2x) = 10
x <6
x, 20
The optimum solution (obtained by TORA) is x, = 0, x, = 5 with
P, = 40, which shows that the most exposure we can get is 40 million
persons.
Step 2. Add the constraint 4x, + 8x, = 40 to ensure that goal G; is not degraded.
Thus, we solve LP, as
Minimize P, = 8x, + 24x,358
Chapter 8 Goal Programming
subject to
x, + 2x, = 10
x =6
4x, + 8x, = 40 (Additional constraint)
x, 20
The TORA optimum solution of LP, is P; = $96,000, x, = 6 minutes, and x, =
2 minutes, It yields the same exposure (P, = 40 million people) but at a smaller cost
than the one in Example 8.2-2 where the main objective is to satisfy rather than opti-
mize the goals. >
‘The same problem is solved now by using the column-dropping rule. The rule calls
for carrying the objective rows associated with all the goals in the simplex tableau.
LP, (Exposure Maximization): The LP, simplex tableau carries both objective rows,
P, and P,, The optimality condition applies to the P,-objective row only. The P,-row
plays a passive role in LP,, but must be updated with the rest of the simplex tableau in
preparation for the optimization of LP,
LP, is solved in two iterations as follows:
Iteration Basic. =) —_Solution
1 Fe Se 8 0
P “8 -24 00 0
31 T 2 1 0 70
% 1 oo 6
2 Pi oo 4 0 40
P; 4 o 2 0 120
x } 1 3 0 5
% i o oO 1 6
‘The last tableau yields the optithal solution x, = 0, x) = 5,and P, = 40.
The column-dropping rule calls for eliminating any nonbasic variable x, with
z, - ¢ ¥ 0 from the optimum tableau of LP, before LP, is optimized. The reason for
doing so is that these variables, if left unchecked, could become positive in lower prior-
ity optimization problems, which would degrade the quality of higher priority solutions.
LP, (Cost Minimization): The column-dropping rule eliminates s, (with z, ~ ¢, = 4).
We can see from the P,-row that if s; is not eliminated, it will be the entering variable at
the start of the P,-iterations and will yield the optimum solution x; = +) 0, which will
degrade the optimum objective value of the P,-problem from P, = 40 to P; = 0. (Try it!)
‘The P,-problem is of the minimization type. Following the elimination of s), the
variable x, with z,— ¢ = 4(>0) can improve the value of P;. The following table
shows the LP, iterations. The elements of P,-row has been deleted because the row no
longer serves a purpose in the optimization of LP;