模糊控制算法分析及改进

基于E.H.Mamdai等人的模糊控制算法,本文讨论了模糊控制器参数的选择及对系统输出性能的影响。与已有的方法不同,本文注重物理概念的分析,并研究了模糊控制器参数的自调整问题。这是因为基本模糊控制器参数的选择难以同时兼顾控制系统的动静态性能,一般只好采取折衷的方法。为了改善控制系统性能,一种实用的方法是在线调整模糊控制器的参数。文中提出了一种分段式自调参方案,仿真结果表明该方案优于已有的自调参方案。为了达到有效的调参目的,本文给出了模糊控制器参数初始配置的一种关系,并通过仿真,验证了该关系是确切的。最后,基于模糊控制算法的特点,探讨了一种新的模糊控制算法的构思,作为今后进一步研究的问题。

Analysis and Improvement of the Fuzzy Control Algorithm

This paper deals with the choice of the parameters of the Fuzzy controller and the corresponding influence upon the performance of the control system based on the Fuzzy control algorithm proposed by E.H.Mamdani and his colleagues. It is different from the existing analyses about the fuzzy control that physical explanations are emphasized in the paper. Furthermore, the problem of the self-tuning of the parameters of the fuzzy controller is studied. Because it is difficult that the control system satisfies the performance of both dynamic and static states at the same time by choice of the parameters of the fuzzy controller, a compromised choice is usually done. For improving the performance of the control sytem , a useful technique is to regulate the parameters of the fuzzy controller on-line. Therefore, a divided-stage self-tuning method is presented in the paper. The simulation result shows the proposed method is superior to the existing one. At the same time,an initial assignment relation about the parameters of the fuzzy controller is given so that the good effect of the regulation can be achieved. The relation is proved to be exact. Finally, a new consideration of the fuzzy control algorithm , which may be taken as a further research subject, is explained in terms of the features of the basic fuzzy control algorithm.