基于参数优化的一般解模糊方法  

Generalized Defuzzification Method Based on Parameter Optimality

作　　者：姚俭[1] 吴健中[1] 徐福缘[2] 

机构地区：[1]上海交通大学管理学院,上海200030 [2]上海理工大学系统科学与系统工程学院,上海200093

出　　处：《系统工程理论方法应用》1998年第3期1-5,共5页Systems Engineering Theory·Methodology·Applications

摘　　要：提出了最优解模糊原理 ,并将这一原理在连续映射的线性赋范空间中公式化。应用逼近原理将求最优解模糊映射问题转化为求参数最优问题 ,提出了 POD这样一个具体的实现方法。运用 Kalman滤波技术完成一定优化指标下的最优参数估计问题 。This article proposes the Optimality Principle of Defuzzification.With this principle,the defuzzification process is designed as mapping in a linear normed space so as to optimize the performance index.The approximation theory is applied to help converting the problem of finding the optimal defuzzification mapping to the one of finding a set of optimal parameters for a given form of functions.As a realization of above work,this article propose the POD method.The POD method provides optimal defuzzification through their parameter learning processes using Kalman filter.Finally the simulation results are given which show advantages of the POD method.

关 键 词：解模糊 参数优化 POD方法 模糊控制 卡尔曼滤波 

分 类 号：O231[理学—运筹学与控制论] O159[理学—数学]