A Design Method of the Optimal PID Controller for Multiple Objective Optimization Using a Genetic Algorithms

Abstract

In this paper, we propose a method of Genetic Algorithms (GA) for multiple objective optimization problems. And, it is shown that the proposed GA improve the convergence of GA and the diversity of Pareto-optimal solutions, through two kinds of easy example problems of 2 objective optimization. Then, we apply the proposed GA to PID control design of 2 objective optimization problems. We solve 2 disk type mixed sensitivity problem and robust sensitivity minimization problem that include limiting control variable problem. The effectiveness of the PID controllers obtained by the proposed method is confirmed by the characteristics of the step response and gain diagrems of sensitivity function and complementary sensitivity function.