利用Nelder-Mead单纯形法优化模糊逻辑控制器

Optimizing Fuzzy Logic Controller Using Nelder and Mead’s Simplex Algorithm

中文关键词  模糊逻辑控制  非线性系统  优化技术

英文关键词  fuzzy logic control  nonlinear system  optimization technique

中文摘要

      本文提出了一种优化模糊逻辑控制器的新方法。该方法的主要思想是自动地优化误差变化率的隶属度函数，因为这类隶属度函数表征了速度的反馈。为此，首先定义了一族参数化的隶属度函数。然后，在系统的运行过程中利用Nelder-Mead单纯算法来优化这类隶属度函数。为了验证所提方法的有效性，报告了控制一非线性被控对象的阶跃与跟踪响应。

英文摘要

      This paper presents a new method for optimizing a fuzzy logic controller. The main idea of this study is to automatically modify the membership functions regarding change-in-error e, which represent the feedback of velocity. For this purpose, first of all, a set of membership functions are chosen by inverstigating their effects on control performance. Then, the Nelder and Mead’s simplex algorithm is adopted to optimize these membership functions during operation. In order to demonstrate the effectiveness of the proposed method, simulation results are reported involving both step and tracking control of a nonlinear plant.