一种基于类覆盖和粒子群优化的模糊神经网络系统  

A Fuzzy Neural Network System Based on the Class Cover and the Particle Swarm Optimization

作　　者：黄艳新[1] 周春光[1] 邹淑雪[1] 王岩[1] 

机构地区：[1]吉林大学计算机科学与技术学院,长春130012

出　　处：《计算机研究与发展》2004年第7期1053-1061,共9页Journal of Computer Research and Development

基　　金：国家自然科学基金项目 ( 60 175 0 2 4) ;教育部"符号计算与知识工程"重点实验室基金项目

摘　　要：提出一种基于类覆盖获取有向图和粒子群优化方法的模糊神经网络模式识别系统模型 ,该模型利用改进的贪心算法获得半径较均匀的超球体类覆盖 ,再利用超球体类覆盖实现模糊输入空间划分和模糊IF THEN规则提取 ,以此实现模糊神经网络系统的结构辨识 ;采用改进的模糊加权型Mamdani推理法确定系统的输出 ,并使用基于粒子群优化的算法对系统参数进行精炼 ,使系统具有很好的强壮性和识别率 对A fuzzy neural network identification model is developed based on the class cover catch digraph and the particle swarm optimazition method. As structure learning of the fuzzy neural network,an improved greedy algorithm is presented for getting the hypersphere class covers with relatively even radii,which is used to partition fuzzy input space and extract fuzzy IF-THEN rules. A weighted Mamdani inference mechanism is improved for the system output and a particle swarm optimization(PSO)-based algorithm is used for optimizing system parameters that improve the system's correct classification percentages and robustness. Experimental results show that the system is feasible and effective to identify 11 kinds of mineral waters by its taste signals.

关 键 词：类覆盖问题 类覆盖获取有向图 粒子群优化 模糊神经网络 模糊输入空间划分 贪心算法 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程] TP391.41[自动化与计算机技术—控制科学与工程]