Fuzzy rules objective function and its feasible solutions in model optimization
Ahmad Lotfi · Nottingham Trent University's Institutional Repository (Nottingham Trent Repository) · 2001
ABSTRACT: In this paper a comparative study of optimization techniques for an optimization problem witha set of linguistic fuzzy rules as the objective function is presented. Two main categories of proposed solutionsnaming gradient methods and direct search methods are explained.KEYWORDS: Fuzzy Rule-Based System, Optimization, Objective Function, Gradient, Direct Search INTRODUCTION The first article introducing the concept of fuzzy constraints and fuzzy objective was published by Bellmanand Zadeh [4]. Since this original work, fuzzy optimization and fuzzy (linear, stochastic, integer, dynamic, ...)programming have received the attention of many researchers [1, 12, 14].In many optimization (decision making) problem we are processing imprecision, incomplete and vague infor-mation [3]. The optimization problem stated here is a method for solving a class of nonlinear programmingproblem. It is used to optimize (minimize/maximize) a nonlinear objective functions which is in the form offuzzy rule-base systems (FRBSs) subject to the constraints. Fuzzy rule-base programming deals with situationswhere variables of optimization are fuzzy variables defined in the universe of discourse specified by the boundaryconstraints.