Yuzo Yamane
Department of Mechanical Engineering Ashikaga Institute of Technology
268-1 Omae, Ashikaga-city, Tochigi-prefecture, 326-0845 Japan
Tel:(81)284-62-0605
Fax: (81)284-62-9802
Xiaojun Zhang
Department of Mechanical Engineering Ashikaga Institute of Technology
268-1 Omae, Ashikaga-city, Tochigi-prefecture, 326-0845 Japan
Tel:(81)284-62-0605
Fax: (81)284-62-9802
This paper deals with the optimal design problem of model output following control in which there are nonlinear disturbance and uncertain parameters, where the output is regulated to follow the output of reference model. In order to determine the optimal feedback and feedforward parameters of the controller, a genetic algorithm is carefully designed. The model output following control is also called an approximate model matching control whose performance and dynamic property are determined by the reference model,that is, the control will correct the plant so as to concide with the property of reference model. A lot of research works on output feedback has been restricted on the area of stability. The research works concerned with model following control have also been done in the control theory of state feedback. When the plant model involves uncert tain parameters (such as unknown disturbance, nonlinear term), optimal control, robust control as well as other linear control systems are discussed in a wide range. However, the discussed techniques of output feedback are focused on the stability of control system not but the design of reference model output following control system. The purpose of this paper is to provide a genetic algorithm for optimal design of gain parameters in a reference model output following control system in which the output of plant model is regulated to follow the output of reference model, based on plant state feedback and reference model state feedforward. Furthermore, some numerical examples are presented to illustrate the effectiveness of the proposed genetic algorithm. Keywords: genetic algorithm, model following control, uncertain systems.