区间二型TSK模糊逻辑系统的混合学习算法的研究  

Research on Hybrid Learning Algorithm Based on Interval Type-2 TSK Fuzzy Logic System

作　　者：郭小蕾[1] 范秋枫 张智峰[1] 王涛[1] 

机构地区：[1]辽宁工业大学电子与信息工程学院,辽宁锦州121001

出　　处：《辽宁工业大学学报（自然科学版）》2016年第3期150-155,共6页Journal of Liaoning University of Technology(Natural Science Edition)

摘　　要：基于A1-C1型区间二型TSK模糊逻辑系统的设计算法展开研究,将模糊聚类、神经网络、模糊逻辑系统相结合设计系统,首先应用模糊K均值聚类方法筛选规则,其次将模糊逻辑系统融入神经网络中设计6层模糊神经网络系统,最后通过合并最小二乘法和BP算法来调节系统前后件参数,使系统取得更好的性能。This article starts the research based on design algorithm of A1 - C 1 interval type 2 TSK fuzzy logic system, and designs a system by combining the fuzzy clustering, neural network and fuzzy logic system. Firstly, fuzzy k-means clustering method is applied to filtering rules; Secondly, the fuzzy logic system is blended with the neural network to design a six-layer fuzzy neural network system; Finally, least square method and BP algorithm are applied to adjust the system antecedent parameters and consequent parameters. It can make the system achieve better performance.

关 键 词：二型TSK模糊逻辑系统 神经网络 模糊K-均值聚类 最小二乘法 BP算法 

分 类 号：O159[理学—数学]