Multiple Objective Programming

Optimization Algorithms for Multiple Objective Systems

Abstract

Optimization algorithms for multiple objective systems, in which there exist plural independent objective functions, are studied. Several algorithms for the unique solution to the multiple objective programs are proposed by introducing the concepts of satisfaction condition and the preference order among objective functions and by extending nonlinear programming procedures (the method of centers, the method of feasible directions). The solutions obtained are proved to satisfy Pareto optimality, which is no doubt a basic criterion, although it gives merely conservative rationality for vector minimization problems. The purpose of this paper is to develop new algorithms for the multiple objective programs.