Summary: | 碩士 === 淡江大學 === 電機工程學系 === 85 === It is necessary to control a system under a stable and safe state. Besides sufficiently realizing the characteristics of the system, the design of a controller is also a key point. That will improve the quality of the controller and meet with user''s requirements. Many approaches have been presented to design controllers in literatures. But most of them are often complicated. In this paper, an effective and simple algorithm, called the two-stage genetic algorithm (TSGA), is presented. It is based on the genetic algorithm and fuzzy set theory. This proposed algorithm has the following characteristics:
(1) The model of the plant to be controlled is not required.
(2) It is simple and easy to implement.
(3) The operator''s experience can be adopted into fuzzy rules and membership function to improve the performance of the controller.
(4) It has the ability to handle the multi-objective control problem and meet the operator''s requirements.
(5) It allows the designer to find an acceptable non-inferior solution.
From simulation results, the performance of TSGA fuzzy controller can be further improved than that of traditional fuzzy controllers. Also, it is expected that the proposed approach can be applied to other applications with good performance.
|