## Are you sure?

This action might not be possible to undo. Are you sure you want to continue?

ELSEVIER

European Journal of Operational Research 83 (1995) 581-593

Theory and Methodology

**Three models of fuzzy integer linear programming
**

F. Herrera *, J.L. Verdegay

Department of Computer Science and Artificial Intelligence, University of Granada, 18071 Granada, Spain

Received May 1992; revised June 1993

Abstract In this paper we study some models for dealing with Fuzzy Integer Linear Programming problems which have a certain lack of precision of a vague nature in their formulation. We present methods to solve them with either fuzzy constraints, or fuzzy numbers in the objective function or fuzzy numbers defining the set of constraints. These methods are based on the representation theorem and on fuzzy number ranking methods:

Keywords: Integer Programming; Fuzzy constraints; Fuzzy numbers

1. Introduction Integer Linear Programming (ILP) problems have an outstanding relevance in many fields, such as those related to artificial intelligence, operations research, etc. They are especially important for representing and reasoning with propositional knowledge. Thus, the use of Mathematical Programming (MP) techniques for treating propositional logic is useful. In particular, several research efforts have involved the use of MP as a tool for modeling and performing deductive reasoning. A n arbitrary system of rules can be represented and solved as an Integer Linear Program. The applications of Integer Programming to logic lead to new algorithms for inference in Knowledge-Based-Systems [13,15,20,23]. A classical ILP problem can be written as follows:

max z=cx

(1)

i ~ M = { 1.... , m},

s.t.

~ a i j x j <_bi, yeN xi>_O , xj~N,

j~N={1

j~N,

. . . . . n},

where N is the set of integer numbers, c e ~n and aij, b i ~ ~, i ~ M , j ~ N .

* This research has been supported by DGICYT PB92-933. * Corresponding author. 0377-2217/95/$09.50 © 1995 Elsevier Science B.V. All rights reserved SSDI 0377-2217(93)E0338-X

Verdegay/ European Journal of Operational Research 83 (1995) 581-593 In real situations however the information available in the system under consideration not of an exact nature. i ~ M. Consider a linear membership function for the i-th constraint. FILP problems with fuzzy constraints This problem can be written as Max s. ifbi<aix<bi+di.+ (0. that is. along with a description of each of the possible problems. j~N (2) i~M.+di)-a._. and an initial study of the fuzzy boolean linear programming problems with fuzzy constraint was carried out. In [12] a classification of them was shown. j~N.582 F. . J. where some lack of precision of a vague nature may be assumed on their formulations. In the following section we will firstly discuss the FILP problem with fuzzy constraints. iEM. i /-L/(X) = (b. a modified objective function and some supplementary constraints and variables. following the ideas expressed ~n [18. This problem was studied in [8]. In view of this classification we will study the FILP models to ascertain whether there are either fuzzy constraints. xj~N.L. The aim of this paper is to study different problems. some conclusions will be pointed out. where an auxiliary ILP problem was presented as a transformation of t h e former FILP problem into a deterministic model with linear constraints.12] we will present an alternative model which allows a fuzzy solution of the problem to be obtained according to the use of the representation theorem of fuzzy sets. (3) Each of these gives the degree to which each x ~ ~n accomplishes the respective constraint. The symbol < means that the decision-maker is willing to permit some violations in the accomplishment of the constraints. he considers fuzzy constraints defined by membership functions ]'~i: ~ n . i f a i x >_b i + di. z = cx ~ aijx j < b i . j~N. Herrera. providing a tool helping reasoning in imprecise Knowledge-Based-Systems. and denote for each constraint Xi={x~nlaix~bi. Next.t. then (2) can be rewritten as (4) max { z = cx I x E X } . This kind of problems will be called fuzzy integer linear programming (FILP) problems. or fuzzy numbers as coefficients in the objective function or fuzzy numbers defining the set of constraints. xj~}.1].x]/di if aix <_bi. 2. in section three the FILP problem with fuzzy numbers in the objective function and in section four the FILP problem with fuzzy numbers in the technological matrix. If X = Iq r ~ MSi xj>O. x]>0. Finally.

In [7] a method was presented which allows us to obtain the fuzzy solution to a fuzzy linear programming problem with nonlinear membership functions from the fuzzy solution associated to the fuzzy linear programming problem with linear membership functions and same right margins [7..L. S ( a ) = { x ~ ~ " [ c x = m a x cy. z = 2x~ + 5 x 2 2 x l . X(a)= ['7 { x ~ n l a i x < . The initial problem (2) may be presented with nonlinear membership functions for the constraints. x j > . denoting Va ~ (0. the fuzzy set defined by the membership function ( sup a x~ U S(a). i ~ M}. i~M. xy E ~ } i~M with ri(a) = b i + di(1 .t. (6) i~M. J. y ~ X ( a ) } . Va c (0. Therefore. In this way. an a-cut of the constraint set will be the classical set X ( a ) = {x ~ R ~ I tZx(X) > a} where V x ~ ~n. X i ( a ) will denote an a-cut of the i-th constraint. to elsewhere. As was shown in [7].F. Then.x2 % 9. s. iXx(X) = inf{/zi(x). As Va ~ (0. ae(O. x i >_ O. the use of nonlinear membership functions do not interfere in the computational efficiency of the solution method. In [9] it was shown that in all cases the objective function associated to the fuzzy solution is included into the same interval.a).r i ( a ) .292-293. In [1] an approach was shown to solve (6).. . xj~l. Proposition 2]. the form of the membership functions does not make the use of the representation theorem complicated. 2 x I + 8x 2 < 31.1]. Herrera.1. (3) can be written as the following auxiliary parametric ILP problem: max z = cx aix<bi+di(1-a). Numerical example Consider the following problem: max s.1]. is the fuzzy solution of the problem (2) (Orlovski.O . j~N. These results may be applied directly to the FILP problems. Xl~ X2 ~ ~ .t. Verdegay~European Journal of OperationalResearch 83 (1995) 581-593 583 It is clear. [16]). 2. X2 >___ O.1]. and by means of the parametric solution of (6) the fuzzy solution to (2) is obtained with a membership function like (5). Xl.1].

Hence. From this point of view two ways to solve (7) can be approached.25. X 2 :> 0 . J. Verdegay/ European Journal of Operational Research 83 (1995) 581-593 with d 1 = 3 and d 2 = 4 the right margins allowed by the decision-maker. 0<a_<l.75. XjE~.584 F.a ) . and finally. in order to solve the optimization problem. Va ~ (0. j~N. F ( ~ ) being the set of real fuzzy numbers.3)/1}. (8) expressing the lack of precision on the values of the coefficients that the decision-maker has. and the costs in the objective are fuzzy numbers. For each feasible solution. there is a fuzzy number which is obtained by means of the fuzzy objective function.75]. Z= ~_~ cjxj j~N (7) ~ aijx j < b i . X 2 ~ [~. The problem can be written as max s. 3.L. obtaining both the optimal solution and the corresponding fuzzy value of the objective. i ~ M.t. Applying the resolution method the following solution is obtained: x ( a ) = (5.a ) .3)/0. j ~ N. 0.3). z ( a ) = 25 z ( a ) = 23 z ( a ) = 21 Va ~ (0. The auxiliary parametric integer programming problem is max s. 2 x 1 + 8x 2_< 31 + 4 ( 1 .3)/0.x 2 < 9 + 3 ( 1 .t.25]. 0. methods ranking the fuzzy numbers obtained from this function may be considered. the fuzzy solution is the following fuzzy set: S = {(5. Va ~ (0. where aij.75. 2x 1 + 5x 2 2x 1 . j~N. xj>O. each of which will provide a different auxiliary conventional optimization model solving the former problem. X 1. with coefficients defined by fuzzy numbers.25. that is. . 1]. that is. The second approach will explore the behavior of the representation theorem for fuzzy sets when it is used as tool to solve the proposed problem. (4. b i ~ ~ are real coefficients. Thus. FILP problems with imprecise costs H e r e we study FILP problems with imprecise coefficients in the objective function. The first will consist of the use of several well known ranking fuzzy numbers methods. x ( a ) = (3. j~N. (3. cj ~ F(~).1].3). x ( a ) = (4. one has the membership functions /zj:~---> [0. X l . j~N i~M. Herrera.3).

. cj. Verdegay/ EuropeanJournalof OperationalResearch83 (1995)581-593 3. then f ( A ) > f ( B ) . f ( A ) = f(B). j ~ N.L. cx < z < Rx . the preference of the decision-maker. e = ( e l .r. j~N cj~F(~). f is called an L R F if VA. The use of fuzzy number ranking methods 585 In this section.rx) if x > O. Many fuzzy numbers ranking methods (FNRM) can be found for instance in [2] and [10]. This paper will focus on those FNRMs which are defined by means of a ranking function. . Then the membership function of the fuzzy number y is given by z) = ( z . ) is known. Usually. Consider A. This nonfuzzy objectivereflects. f ( A ) < f ( B ) are equivalent to A > B. and xj _> O. If we apply different methods of ranking fuzzy numbers to (7) then it is interesting to observe how the optimal solution to it will be an optimal solution of a conventional Programming problem with similar constraints and a nonfuzzy objective function. . (lO) As it is well known. A simple method of comparison between them consists of the definition of a certain function f : F(R) ~ ~. . then d -x and d' -x will be the lateral margins (right and left respectively) of the fuzzy number cx.1. . which is not too restrictive because many well known F N R M can be formulated by using linear ranking functions in some way. and their membership functions supposed in the form ((u-r~)/(cj-rj) Vu~.J. Vr~R r>O. A = B. and particularly by means of a linear ranking function (LRF). ifc. Herrera. by means of the ranking functions. triangular fuzzy numbers will be considered. (12) =/0j(z ) where r = (r 1. They will be denoted c] = (rj. g:X~F(~). Consider the set of fuzzy numbers A = {g(x)l x ~ X } . tzc. . from this definition several FNRPs may be considered. . To simplify. otherwise. B ~ F(~). . If this function f ( . if x > O. otherwise. the problem now is how to determine the greatest in A. . rx < z < cx. A < B respectively. In [3] an extensive study of these L R F can be found. B e F ( ~ ) . . and g a function mapping the set of feasible alternatives of (7) into the set of fuzzy numbers. e n) and R = (R 1.rx)/(cx . Ri). j~N. (9) where extended sum and product by positive real numbers have been considered defined in F(R) by means of the Zadeh's Extension Principle. Let us assume a linear expression y = Ejcjxj in which the c / s are fuzzy numbers with membership functions similar to the ones given by (1]).c and d' = c .(u)=~(Ry-u)/(Ry-cy) ifrj<u<cj. g(x)=cx= ~] cjxi. The problem of comparison of fuzzy numbers has been widely investigated in publications. . let X be the set of feasible alternatives of (7). rn) .<_u<_Ri. Then x* ~ X will be said to be an optimal alternative if the fuzzy number g(x*) is greatest in A. (11) to T h e n the following result holds [11].F. Rn). If it is denoted that d = R . . f(A+B)=f(A)+f(B)and f(rA)=rf(A).. Hence. .

t.)]...a)]. xj e ~1}.a ¢~hj-l(1 . max {[c + ( d . 3. x i e ~} which is nonlinear.. ~ .a } . x e X } . The use of the representation theorem Consider Vc e R n.1]} (21) . xj e ~}. (13) xje~ .)] and qt = [~Fx(. J It is clear.). ) = [~ba(. c = (ca.d ' ) / 4 ] x [Ax _<b. J from (12).~(c) > 1 .(.22] produces respectively the models max {[c + (d .. a ~ [0.). the membership function ~ ( c ) = in. ~b... as was shown in [19]. Moreover. and denoting ¢i -= h71. j e N.d ' x ) / 6 ] A x _<b. taking into account that t~(c)>l-ac*inf~j(cj)>l-ac~. (3cx + dx ..a ) < c j < g i ' ( 1 . denotes the set of vectors c e •" with all of their components ei in the interval [~b/(1 . different auxiliary models solving (7) can be obtained.L. (16) (17) (18) (15) max {(cx + d x ) / ( d x + 1) lAx < b. Clearly.j(cj)>l-a.1]} where ~ ( .a). f(cx) Ax < b .a). .f ixi(cj) .d')/3]x lax <_b. second and third Index of Yager [21. ~j(1 .ae[0. b) The use of first.g y -1.a ) ... J e N. Consequently a fuzzy solution to (7) can be found from the solution of the Multiobjective Parametric Integer Linear Programming problem max {cx ]Vc e ]~":/. jeW. q)(1-a)<c<~(1-a). xj e ~}. if we use L R F then the auxiliary problem obtained in (13) will be the following ILP problem: max ( Y] f(c])xi[j e N . ~ j~N (14) Some of these auxiliary models are shown in the following a) The use of the Index of Chang [4] provides the problem max {(dx + d'x). that ~(-) defines a fuzzy objective which induces a fuzzy preorder in X.j-(cj-) > 1 . j e N .1].a ) . jeW.1]. (20) is obtained. a e [0. (20) can be finally rewritten as max {cx]x e X ..586 F. ae[O. the problem (19) can be written as max{cx[x~X. xEX (19) But. Verdegay /European Journal of Operational Research 83 (1995) 581-593 Consider a ranking function f mapping each fuzzy set into the real line. Then a solution for (7) can be found from max s. if F(1 .. jeN. J. Therefore according to the ranking function f used. cn). c e F ( 1 .. f : A --* R. qt (.2. Herrera.

T h e n conventional parametric LP problems are obtained. + r i . x2nx) n.2 a~[0. one can define s = U ot s(1 - . If the set of optimal points of these is defined as S~(1 .a ) is the subset constituted by vectors whose j-th component is equal to either the u p p e r or the lower bound of ci. according to some results by Ishibuchi and Tanaka.a) =1 ~ [ 2 c i + a ( R .a ) is defined as the set of solutions of the auxiliary problem considered according to the two approaches. Verdegay~EuropeanJournalof OperationalResearch83 (1995)581-593 587 which for each a ~ [0.r i ) ] x j and zC(x. . P(a): max s. (23) xi~. and ~0i = 0 for 0 < i < t.J s ~ ( 1 .1] is a Multiobjective ILP problem denoted M(a) and having in its objective function costs that can assume values in the respective intervals. for every a ~ [0. (22) A x _<b. C2X . ~bj(1 .(1 . zC(x.. °Jr = 7v2. in which S(1 . the resolution of all problems in the family {M(a). o~). j~N. . the fuzzy solUtion for (7) can be found from the parametric solution of the following biobjective parametric problem. Herrera..1] to each of the objectives taking part in (22) or (23). ) ] x i . It is enough to 1 1 take to1 = (Vx + ~92). ..J. a ) = E [ c i . j ~ N.1]} where the fuzzy solution for (7) will be obtained from the solution of the following Multiobjective Integer Programming problem: max s.F. (22) or (23).a ( c j .1].2 c . . [14].a ) o F ( 1 . (clx. j=l j=l Now. x>O. z ' ( a ) = ( z l ( x .) which is a fuzzy set giving the fuzzy solution to the former problem. tot) and 13 = u = (91. following [5].a ) or ~. . where zl(x. .t. such that Ek/3k = 1. on the use of interval arithmetic for solving LP problems with interval objective functions. .. and the equivalence is obtained. .. in accordance with the representation t h e o r e m for fuzzy sets.1].L. ~ [0. Let us assume (22) and (23) and consider /3 = t o = (w 1. ck~E(1--a).2. Different alternatives can be considered here: first. a ) ) Ax < b . then the fuzzy solution with weight/3 will be given by the fuzzy set S~ = [.1]. a) in the case of triangular fuzzy numbers are defined by n z l ( x . Concretely. . . On t h e other hand. k=1. a decision-maker may be able to assign weights/~k ~ [0.t.1].a ) for every a ~ [0. where E(1 . a ~ [0.a ) . o~ In [11] it was shown that using weight vectors then P~(a) is a particular case of M s ( a ) . u 2) then these problems are denoted Ms(a) and P~(a) respectively. a) and zC(x.a).

the optimal solution of which is x ( a ) = (7. 2 x 1 q. 2 x I q. j~N. . xj~N. Verdegay~EuropeanJournalof OperationalResearch83 (1995)581-593 3. j~N.875].875.a ) = 3 . 0.a ) = 3 + 2 a .1]. xj>O.1]. ( 5 . Herrera. Xi>__O.1].25]. a~[0. Va~[0. and the associated interval p a r a m e t r i c p r o b l e m is: max s. x ( a ) = (1. W e have the following functions: ~bi(1 . (3 + 2 a ) x 1 + 5x2} 2 x I .t.25.2). 4 ) / 1 } . F r o m (23) the auxiliary multiobjective I L P p r o b l e m is: max s. 8 7 5 .10a z(a)=23-2a 'Ca ~ [0.2 a . j~N.4). z = C l X 1 "1. xj>_O.3). 3.x 2 < 12.x z < 12. 0.t. ( 1 . 2 ) / 0 .t. 2 x 1 + 8X 2 ~ 35. z = (3 . 1/t1(1 .t.3.St3 = { ( 7 .0) a n d / 3 = (0.1]. .2 a ) x l +5x2.8X 2 ~___ 35. Next.5.5) F o r / 3 = (1. j~N.5 X 2 2 x 1 . Numerical example Consider the following problem: max s. Xj~N. Z=ClX ~ + 5x 2 2 x 1 .x 2 < 12.2 a ) x 1 + 5 x 2 2 x 1 --X 2 _~ 12.0) the auxiliary p a r a m e t r i c p r o b l e m is: max s. xj>O. = (5. Va ~ [0. 3 ) / 0 .2a <c 1 < 3 + 2a. 2 5 . ~ [0. 2 x 1 + 8X 2 ~ 35. we solve the above auxiliary multiobjective I L P p r o b l e m for the following weight vectors: /3 = (1. 3.8X 2 ~ 35. { ( 3 . 5).588 F. Xy~N. a~[O.L.J.0. xj~N. w h e r e c = (1. z(a)=31-14a z(a) = 30.

x j ~ ~. Using Ranking Function the auxiliary problem obtained is a) Using the ranking function of Chang. 2x 1 q. which contains good alternatives. and hence the decision maker eventually makes the final choice himself. S. aiix i < b i.. max {3x 1 -t.x 2 < 2 12. xi>O.2).5X 212x I .L. 3~iEF(R ) s u c h t h a t tzi:R---> [0.x 2 < 12. j ~ N . As we can see according to the solution method used we have different solutions. 2x I + 8x 2 < 35. xj > O. max {2Xl + 3. m a x { 5 X l + -I-X 2 1 2 x 1 . xj>O.t.2). J. and the corresponding optimal solutions are x* = (7.x 2 < 1 2 . j~N. max {3x I + 5x 212x I . z=cx ~. the optimal solution of which is x* = (7. j ~ N } . the following membership functions are considered: For each row (constraint) in (24). xj~[~.5): max s. Xi > O. x j > 0. 2x 1 + 8x 2_<35. When the representation theorem is used then a fuzzy solution is obtained.F. b i ~ F(E). Xj ~ N. Remark.t. j ~ N } . b) Using the ranking functions of Yager.2). (26) . 2 ) / 1 } . z=3x 1+5x 2 2xl .1]. For each i ~ N and j ~ N. . that the decision-maker permits certain flexible accomplishment for the constraints. j ~ N .1]. j ~ N}.5.Lij : ~ "'> [0. :~]'~ij E~_ F(R) such that ]. i~M. 4.33x~x 212x 1 . Verdegay~European Journal of Operational Research 83 (1995) 581-593 589 F o r / 3 = (0.1]. which is in accordance with the imprecise raising of the problem.x 2 < 12. j~N (24) i ~ M.4) and x* = (7. . as in (2).8X 2 _~<35.0. 2x I + 8x 2 < 35. The symbol < means. where aij. xy>_O. x ( a ) : (7. and the corresponding optimal solution z ( a ) = 31 Va ~ [0. Herrera. a~[0. x* = (0. 15 1 . (25) which defines the fuzzy number on the right-hand side.2) respectively. x j ~ N .1]. we consider FILP problems with fuzzy numbers defining the set of constraints. = { ( 7 . These can be formulated as follows: max s. xj ~ ~. . FILP problems with fuzzy numbers as coefficients of the technological matrix Now. Thus.x 2 < 12. 2 x l + 8 x z < 3 5 .

L e t f i be a fuzzy number. the accomplishment degree of the fuzzy number with respect to the i-th constraint. X1. (0.t.a ) . giving his allowed maximum violation in the accomplishment of the i-th constraint. xj>_O.t. ]~N.1] iEM.l x2 _< 9. j=l i~M. Then as an auxiliary problem to solve (24). j~N. j~N. 2Xl + 8x2 _<31. 4. Considering this relation as a ranking function.t. a~(0. . (27) giving. fixed by the decision maker. 1 ] . i~N. then (28) becomes the following parametric LP problem: n max s. ~ cjxi j=l n (29) Ef(aij)xj<_f(bi ) + f ( t i ) ( 1 . d: F(R) ailxl +qi2x2+ . cx f i aijx ] (28) <f(bi+ti(1-v~)) . different models of conventional LP problems are obtained. x]>__0. +ainXn. Moreover. according to the characteristics of the relation [-<-]. a ~ ( 0 . X l . . X 2 E [~. For each row of (24). for every x ~ R n. ~ cix j j=l n (28) Eaijxj[~bi-I-ti(1-a). one can propose the following one: n max s. the adequacy between this fuzzy number and the corresponding one bi with respect to the i-th constraint. Verdegay ~European Journal of Operational Research 83 (1995) 581-593 defining the fuzzy numbers in the technological matrix.1. J. that is. Herrera. i~M. the auxiliary model obtained is the following: max s. a~(O. . X 2 >___0 . if we use LRF.590 F.t. j=l x/>O.1]. and. z = 2Xl +5x2 2 x l .1]. where ~ represents a relation between fuzzy numbers. Numerical example Consider the following problem: max s.L.

z = 2 x 1 + 5 x 2. X 2 ~ N .5. (6. b I = (7. 6). t 2 = (3. x 2 ~ N . .a ) .166x2) < [6 + 3(1 . 1. z = 2 X 1 q. X l .1]. J.0. a12 = (0. (4. we apply a linear ranking function.666x 1 + 4.5). ( 2 x I + 1 . Herrera.683. X1.1. 10). 8. 2. 4.923.F. z=2x 1+5x 2 2x 1 !x2~]9 -I- 3(1 . which makes its solution m o r e complicated.t.L. the auxiliary parametric models which represent the preferences according to the ranking method are obtained.a ) ] .583x2) < [3 + 1. 31.333 + 1. I f we use nonlinear ranking functions then nonlinear parametric programming problems are obtained. 2xl + 8x2 Xl.5 x 2 2 x 1 -.583(1 . X 2 ~ 0. max s. 2).2)/0.2)/0. Some approaches based on the representation theorem and on F N R P have b e e n provided to solve them. Verdegay/ European Journal of Operational Research 83 (1995) 581-593 591 with all = (1.1].) and q21 = (1. X 2 (~ ~ .052. 5 x 2 ) ( x 1 .666 + 3. 2. and applying the ranking function for fuzzy numbers. X1.666 + 4. X 2 ~__ O. For example. 3). ( 2 x I + 3x2)(1. or fuzzy numbers in the objective function or fuzzy numbers defining the set of constraints.333(1 -.3)/0. whose optimal solution is S = {(7.2)/1}. 5. (5. 2. 3.a ) . a22 = (7.384. By means of (28) the auxiliary problem is written as max s. a) Using the first index of Yager.t.5. 35).3)/0. 10).a ) ] [15. t 1 = (2.333x I + 8.166(1 -.a ) ] [4. X 2 ~ O.833 + 2. In this example. (3. 3i 4. if we use the index of Chang. ]31 + 4(1 a ~ (0. with either fuzzy constraints.a ) ] . we obtain the next auxiliary problem. max s.t.1]. b) Using the index of Chang.a ) . Conclusions In this p a p e r we study three models for dealing with the lack of precision of a vague nature in the formulation of I L P problems.166x 2 _< 8. xl. a~(0. [4]. a ~ (0. b 2 = (29. X1.166(1 .5.1. the first index of Yager.5(1 . Note how different this solution is from the corresponding one in the parallel model without fuzzy numbers in the technological matrix shown in Section 2. 9.333x 2 _< 31. which is a nasty nonlinear parametric p r o g r a m m i n g problem.

1989. "Imprecise costs in mathematical programming problems". and Vila. Berlin. M. and Vila. M. This p r o b l e m will b e d e a l t with in f u t u r e p a p e r s . Springer-Verlag. 1991. a n d c o n t a i n g o o d alternatives. [11] Herrera..G.. Academic Press. [7] Delgado.. Decision Support Systems 4 (1988) 45-69. North-Holland. "Multiobjective programming in optimization of the interval objective function". Kacprzyk and M.L. Fuzzy Sets and Systems 49 (1992) 1-11. 263-272.A. [6] Delgado.N. Theory.. J. C. t h e u s e o f t h e r e p r e s e n t a t i o n t h e o r e m gives us i n t e g e r p a r a m e t r i c auxiliary p r o b l e m s . L. [17] Taha. and Tanaka. A. [15] Jeroslow. Fuzzy Sets and Systems 29 (1989) 21-29.. M. Logic-Based Decision Support Mixed Integer Model Formulation. 1991. (Amsterdam)..E.D.A. t h e c h o i c e o f a r a n k i n g m e t h o d has effects o n t h e results as w e can see in t h e e x a m p l e s .G. F. [3] Campos. Herrera. and Gillett. Herrera.A. Thesis. Roubens (eds. New York. 1984. J. .. F. [9] Garcia-Aguado. Interactive Fuzzy Optimization... S. w h e n we u s e r a n k i n g m e t h o d s . Conf. Fedrizzi.. [13] Hooker. and Vila.. Journal of the Operational Research Society 31 (1980) 253-262. "Post-optimality analysis on the membership functions of a fuzzy linear programming problem". in: H. Zadeh and B. R. [2] Bortolan. [10] Gonzalez. Gonzalez. M. J.A. on Policy AnaL and Inf.L. W.592 F. Verdegay. 1975. Amsterdam. Fuzzy Sets and Systems 53 (1993) 280-297. [16] Orlovski. "Ranking of fuzzy utilities with triangular membership functions". H.. J. Kybernetes 6 (1977) 197-201. [12] Herrera. University of Granada. Games (eds). J. Fuzzy Sets and Systems 15 (1985) 1-20. and Vila.R. a n d h e n c e t h e d e c i s i o n m a k e r e v e n t u a l l y m a k e s t h e final choice himself.).. "A study of the ranking function approach through mean values". [14] Ishibuchi.. "Fuzzy integer programming". B. [4] Chang. Zimmermann. in: M. F. a n d also t h e d e c i s i o n m a k e r m u s t c h o o s e t h e final r a n k i n g m e t h o d a c c o r d i n g to his p r e f e r e n c e s . A.. M. A l l t h e m o d e l s o b t a i n e d in t h e p a p e r have e x p r e s s i o n s for t r i a n g u l a r fuzzy n u m b e r s . H. Int. M. "A quantitative approach to logical inference". Ph.. w e m u s t also p o i n t o u t t h e n e c e s s i t y o f d e v e l o p i n g a n i n t e r a c t i v e d e c i s i o n s u p p o r t system in fuzzy i n t e g e r p r o g r a m m i n g p r o b l e m s . w h i c h w o u l d allow i n t e l l i g e n t decisions a c c o r d i n g to t h e a c t u a t i o n p r e f e r e n c e s o f t h e d e c i s i o n m a k e r s . "A review of some methods for ranking fuzzy subsets". [5] Delgado.L. t h e solutions o f w h i c h a r e u s e d for b u i l d i n g t h e fuzzy s o l u t i o n o f t h e m o d e l s . Verdegay. [8] Fabian.. Integer Programming. "On programming with fuzzy constraint sets"..A. e a c h m e t h o d has its o w n a d v a n t a g e s over t h e o t h e r s in s o m e p a r t i c u l a r situations. "Approaching fuzzy integer linear programming problems". and Verdegay. European Journal of Operational Research 48 (1990) 219-225.A. Fuzzy Sets and Systems 56 (1993) 47-49.L. b u t all o f t h e m m a y b e easily r e w r i t t e n for t h e c a s e o f fuzzy n u m b e r s o f L R kind.L. Fuzzy Sets and Decision Analysis. O n t h e o t h e r h a n d . J.. Fuzzy Sets and Systems 35 (1990) 29-41. T h e fuzzy s o l u t i o n s a r e in a c c o r d a n c e w i t h t h e i m p r e c i s e raising o f t h e p r o b l e m s .. A s we h a v e a l r e a d y said... "On the sensitivity of membership functions for fuzzy linear programming problems". M. C. Control and Cybernetics 16 (1987) 113-121. J. H. R. 1981... L. North-Holland.. and Degani..L. "Parametric integer programming analysis: A contraction approach". "Problems and algorithms in fuzzy integer programming". G. Applications and Computations. "A general model for fuzzy linear programming". and Verdegay. in: Proc.. Finally.. J. Verdegay /European Journal of Operational Research 83 (1995) 581-593 T h e c o m p u t a t i o n a l b u r d e n o f t h e auxiliary m e t h o d is in r e l a t i o n to t h e n u m b e r o f auxiliary i n t e g e r l i n e a r p r o g r a m m i n g p r o b l e m s t h a t w e m u s t solve.. M. References [1] Bailey.. Systems. "On the use of the ranking function approach to solve fuzzy matrix games in a direct way". 123-131.A. M. and Stoica. Verdegay..J. Acknowledgement W e wish to t h a n k t h e a n o n y m o u s r e f e r e e s for t h e i r v a l u a b l e c o m m e n t s w h i c h h a v e i m p r o v e d t h e presentation of the paper.

Information Science 24 (1981) 143-161. R. R.. 1978. 1978 CDC. Fuzzy Information and Decision Processes. . "A procedure for ordering fuzzy subsets of the unit interval". H. "Fuzzy mathematical programming".L. Herrera..R. "Fuzzy 0-1 linear programs". 1435-1437. M..M. 133-145.A. Sanchez (eds.L. International Journal of Systems Research and Information Science 2 (1987) 81-100. Zimmermann. J.). [22] Yager.R. [21] Yager. "Linear and integer programming applied to the propositional calculus".J. Verdegay / European Journal of Operational Research 83 (1995) 581-593 593 [18] Verdegay. " A dual approach to solve the fuzzy linear programming problem". J. North-Holland.J. in: Proc. Fuzzy Sets and Systems 14 (1984) 131-141.R. in: H.. 231-237..A. Zadeh and B. "Ranking fuzzy subsets over the unit interval". L.R. International Journal of Man-Machine Studies 29 (1988) 685-714.P. "A mathematical programming approach to inference with the capability of implementing default rules". Gupta and E. Amsterdam. [20] Williams. [23] Yager. in: M. [19] Verdegay. North-Holland. Fuzzy Sets and Decision Analysis...L. (Amsterdam).). [24] Zimmermann. H..F. R. and Pollatschek. J. Gaines (eds. 1982. 1984.

- 基于负荷动态模拟的冰蓄冷系统优化控制分析
- lec14.pdf
- Fuzzy Logic
- compressorbhp & cost
- Application of Fuzzy Logic in Operation Management Research
- sun8
- Three Models of Fuzzy Integer Linear Programming
- Cali
- Linear Programming
- SMAW Process Parameters Optimization using Taguchi & Fuzzy Logic
- Week6 Fuzzy Logic 2
- Fuzzy Based Decision Making for Selection of Fluid Film Journal Bearing
- Business Optimization
- Fuzzy Logic 2
- Fuzzy Logic Data
- Management Science Lecture Notes
- Fuzzy Logic Final1
- An Introduction to Fuzzy Control Systems
- Cingolani Alcala-Fdez JFuzzyLogic 2013 IJCIS
- Operation Management-10-NPTEL
- Overview Op Tim Ization
- Modeling
- Rawlings, 2000 - Tutorial Overview of Model Predictive Control
- sl
- IJCST-V2I4P21-Author:M. Mayilvaganan, K.Rajeswari
- 94-471-1-PB
- ZR-04-19
- Combined Fuzzy Logic and Genetic Algorithm Techniquesto an Electromagnetic Field Problem
- Fuzzy Logic Application
- Procedures for Providing Robust Gate Assignments for Arriving

- Solving Fuzzy Matrix Games Defuzzificated by Trapezoidal Parabolic Fuzzy Numbers
- A Comparative Study of Fuzzy Logic with Artificial Neural Network Techniques in Cancer Detection
- A Review and Analysis on Mobile Application Development Processes using Agile Methodologies
- Unification Of Randomized Anomaly In Deception Detection Using Fuzzy Logic Under Uncertainty
- A Survey on Hybrid Soft Computing Techniques based on Classification Problem
- tmpBFF4
- Analysis of CAS Using Improved PSO with Fuzzy Based System
- An Intelligent IR for Natural Language Query Processing using Fuzzy Logic
- tmp5E42
- A Methodology for 4G Mobile Networks Using Vertical Handoff Approach
- A Defect Prediction Model for Software Product based on ANFIS
- Torque Control of BLDC Motor Using ANFIS Controller
- tmp7D96.tmp
- ON Fuzzy Linear Programming Technique Application
- Comparative Analysis of Power System Stabilizer using Artificial Intelligence Techniques
- Power Factor Correction and Speed Control of Luo Converter Fed BLDC Motor Using Neuro Fuzzy
- Improved remote sensing image segmentation using fuzzy logic for dynamic statistical region merging algorithm
- As IEC 61131.7-2004 Programmable Controllers Fuzzy Control Programming
- Development of Adaptive Neuro Fuzzy Inference System for Estimation of Evapotranspiration
- Inverse Kinematics of Robot Manipulator with 5 Degree of Freedom based on ANFIS toolbox
- Design of Fuzzy Logic Controller for Speed Regulation of BLDC motor using MATLAB
- tmp911D.tmp
- Modelling and simulation of power system stabilizer using fuzzy logic
- Fuzzy Logic v. Trapflix and Snopp Dogg amended complaint.pdf
- Correlative Study on the Modeling and Control of Boost Converter using Advanced Controllers
- A Hybrid Approach to Sizing Problem in Software Project Estimation
- Adaptive Fuzzy Filter and Histogram Equivalence In Order to Remove Salt and Pepper Noise
- Performance enhancement of DC motor drive using Fuzzy Logic Controller
- Study of Different Techniques for Load Forecasting-A Review
- Performance of Node Deployment Techniques in WSN

- Comparative study of Conventional and Traditional Controller for Two Equal Area interconnected power plant for Automatic Generation Control
- Emerging Trends of Ant Colony Optimization in Project Scheduling
- UT Dallas Syllabus for opre7330.001.07f taught by Milind Dawande (milind)
- Surveying, Planning and Scheduling For A Hill Road Work at Kalrayan Hills by Adopting ISS Method
- Weight Optimization of Composite Beams by Finite Element Method
- UT Dallas Syllabus for eco5302.501.07f taught by Chetan Dave (cxd053000)
- A Review on Image Segmentation using Clustering and Swarm Optimization Techniques
- Multi-objective Optimization for Ultrasonic Drilling of GFRP Using Grey Relational Analysis
- Determining the shortest path for Travelling Salesman Problem using Nearest Neighbor Algorithm
- UT Dallas Syllabus for opre7313.001.11s taught by Milind Dawande (milind)
- The Optimizing Multiple Travelling Salesman Problem Using Genetic Algorithm
- Comparative Analysis of Optimization Algorithms Based on Hybrid Soft Computing Algorithm
- tmpBFF4
- Analysis of CAS Using Improved PSO with Fuzzy Based System
- Optimization of Water Distribution Network for Dharampeth Area
- Combined Economic and Emission Load Dispatch using PPF using PSO
- A Review on Various Approach for Process Parameter Optimization of Burnishing Process and TAGUCHI Approach for Optimization
- Classification of Lung Diseases Using Optimization Techniques
- tmp5056.tmp
- Optimization Approach for Capacitated Vehicle Routing Problem Using Genetic Algorithm
- tmp5DD5.tmp
- Planning of Supply when lead time uncertain
- Mathematics-III
- UT Dallas Syllabus for math6321.521 06u taught by Janos Turi (turi)
- tmpA166.tmp
- UT Dallas Syllabus for opre7320.001.09s taught by Suresh Sethi (sethi)
- tmp82D3.tmp
- Quantitative Management for Analysis
- Recent Advancements in Scheduling Algorithms in Cloud Computing Environment
- Optimization of Shell and Tube Heat Exchanger

Sign up to vote on this title

UsefulNot usefulClose Dialog## Are you sure?

This action might not be possible to undo. Are you sure you want to continue?

Close Dialog## This title now requires a credit

Use one of your book credits to continue reading from where you left off, or restart the preview.

Loading