In this paper develops a method based on fuzzy logic for solving linear approximation relationship nonlinear optimization problems. In the proposed method, the nonlinear objective function as well as the constrained functions forming the feasible regions is linearly approximated about an initial feasible point by Taylors Series and then the nonlinear optimization problem is approximated to a linear optimization problem or a linear programming problem. The concept of fuzzy decision making is iteratively applied to find the fuzzy optimal solution of the problem.
In this paper, we develop a solution method for solving linear approximation relationship constrained nonlinear optimization problems by using the concept of fuzzy logic. We consider a constrained nonlinear optimization problem. The nonlinear objective function as well as the constrained functions forming the feasible regions is linearly approximated about an initial feasible point by Taylors Series and then the nonlinear optimization problem is approximated to a linear optimization problem or a linear programming problem. The concept of fuzzy decision making [e.g. see Bellman & Zadeh (1970), Warners (1987), Feng (1987), Zhang et.al. (2013), Chamani et.al. (2013)] is iteratively applied to find the fuzzy optimal solution of the problem.
In Sec. 5.2, we consider the procedure for obtaining the initial feasible solution. In Sec. 5.3, the solution method of the problem is investigated. An iteration algorithm is also developed. In Sec. 5.4, a physical numerical example is shown to illustrate the efficiency of the proposed problem.
A. Calculation of Initial Feasible Point
In this section, we consider a method of finding an initial feasible solution of a given system of constraints. A common method used in many practical situations is “trial and error” method [Walsh (1977)].
Now our problem is to construct a fuzzy LP model which is to be solved iteratively.
Before we construct a specific model of linear programming in a fuzzy environment it should have become clear that a fuzzy linear programming is not uniquely defined type of model but that many variations are possible, depending on the assumptions or features of the real situation to be modelled.
First of all we assume that the objective function should reach some aspiration levels which might not even be defined crisply. Thus, we might want to improve the present cost situation considerably and so on.
Secondly, we accept small violation of constraints but also attach different degrees of importance to violations of different constraints.
Here, we shall present a model that is particularly suitable for the type of linear programming model (5.3.4) in fuzzy environment which seem to have some advantages [Werners (1987)]. Werner (1987) suggested the following definition.
 R.E.Bellmen & L.A.Zadeh (1970) “ Decision making in fuzzy environment”, Management Science,17(4),pp-141-164.
 B.Werners(1987), “ Interactive multiple objective programming subject to flexible Constrants”,EJOR 85-PP,342-349.
 X. Fang, H. Zhang & J. Zhou (2013). “ Fast Window Fusion Using Fuzzy Equivalence Relation”, Pattern Recognition Letters, 34, pp. 670-677.
 Y. Feng (1987). “Fuzzy Programming- A New Model of Optimization”, Optimization Models Using Fuzzy Sets and Possibilities Theory (Kacprzyk and Orlovski Ed.), D. Riedel Publishing Comp. Boston, pp. 216-225.
 M.R. Chamani, S. Pourshahabi & F. Sheikholeslam (2013). “ Fuzzy Genetic Algorithm Approach for Optimization of Surge Tanks”, Scientia Iranica, 20(2), pp. 278-285.
 C.V. Negoita & D. Ralescu (1977). “On Fuzzy Optimization”, Kybernates 6, pp. 193-195.