基于神经网络的模糊逻辑设计算法  

Fuzzy Logic Design Algorithms Based on Neural Network

作　　者：李四福[1] 冯伟莉[1] 

机构地区：[1]中国地质大学

出　　处：《武汉汽车工业大学学报》1996年第5期18-22,共5页

基　　金：地矿部"八.五"重点科技项目

摘　　要：提出了一种利用神经网络学习功能来设计模糊逻辑的方法。用类神经网络算法产生模糊规则和隶属函数，通过学习得到的模糊规则、隶属函数和以新模糊推理、逆模糊化为基础的新模糊设计，显著地改进了系统性能，缩短了设计周期，将规则数目和隶属函数最优化。A new design of fuzzy logic neufuz, using neural net learning is proposed. Artificial neural net algorithms are used to generate fuzzy rules and membership functions. Combining learned fuzzy rules, membership functions and a new fuzzy design technique based on a new fuzzy inferencing. Defuzzification significantly improves performance, accuracy, reliability and usage of design time. Neufuz also minimizes system costs by optimizing the number of rules and membership functions.

关 键 词：神经网络 模糊逻辑 算法 设计 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]