Solving the nonlinear dynamic control problems by GA with structurizing the search space

Abstract

We propose a new search method of genetic algorithm (GA), which reduces the difficulties of the design of the fitness function. In the method, the control objective is divided into some intermediate control objectives according to the control strategy. The search process progresses with the fitness function corresponding to the intermediate control objective, and the process is controlled by switching the fitness function based on the average fitness value of the current candidate solutions so that the optimum solution with desired quality may be found. Thus, the search space is structured repeatedly during the search process by switching the fitness function based on the quality of the current candidate solutions. In order to confirm the availability of the proposed method, the swing-up control of the cart-pendulum system is used as an example, and some simulation results are given.