A general approach for solving optimal control problems using optimization techniques

Abstract

A general approach to the determination of approximate solutions of general control problems by exploiting modern global search and optimization techniques is proposed. According to the methodology developed in this paper, controls are represented by discrete vectors and substituted in system equations. The components of these vectors are regarded as variables of a performance index based goal function that is to be minimised with respect to the system constraints. Such an approach enables modeling and solution of a wide class of optimal control problems, arising in engineering practice, within a unified framework of constrained optimization techniques, including implementation of genetic algorithms for global optimization and multiobjective control. Computer realizations of the proposed method are mainly based on MATLAB simulation programs. The results obtained can be implemented to solve optimal control problems in the field of computer aided control engineering, CIM, mechatronics and robotics.