基于最小Wilcoxon学习方法的模糊树模型  

Least Wilcoxon learning method based fuzzy tree model

作　　者：张伟[1] 毛剑琴[1] 

机构地区：[1]北京航空航天大学自动化科学与电气工程学院,北京100191

出　　处：《北京航空航天大学学报》2013年第7期973-977,共5页Journal of Beijing University of Aeronautics and Astronautics

基　　金：国家自然科学基金重点资助项目(91016006;91116002);中央高校基本科研业务费专项资金资助项目

摘　　要：模糊树方法采用最小二乘法学习模糊规则的后件参数,对例外点敏感.为此采用对例外点不敏感的最小Wilcoxon学习方法代替最小二乘法,提出一种基于最小Wilcoxon学习方法的模糊树建模方法,该方法既改善了模糊树方法对例外点敏感的缺点,又继承了模糊树方法的优点.通过对混沌时间序列预测研究,仿真结果表明:所提方法可以对Mackey-Glass混沌时间序列进行准确预测,验证了该方法的有效性和对例外点的鲁棒性.Fuzzy tree （FT） method used the least square method to learn the consequent parameters of the fuzzy rules, so it was sensitive to the outliers. The least Wilcoxon learning method was used to replace the least square method and a robust modeling method against （or insensitive to） outliers was proposed based on the least Wilcoxon learning method, called least Wilcoxon-fuzzy tree （LW-FT）. The proposed method is not only insensitive to the outliers, but also has the advantages of the FT. Finally, the simulations on Mackey- Glass chaotic time series prediction were performed. The results show that the chaotic time series are accurate- ly predicted, which demonstrates the effectiveness and the robustness to the outliers of this method.

关 键 词：模糊树 例外点 最小Wilcoxon学习方法 混沌时间序列 预测 

分 类 号：TP13[自动化与计算机技术—控制理论与控制工程]