智能系统学报Issue(4)：425-431,7.DOI:10.3969/j.issn.1673-4785.201305083
基于 MFOA 和 LW 的混沌时间序列鲁棒模糊预测
Robust fuzzy prediction of the chaotic time series based on the MFOA and LW
摘要
Abstract
Focusing on the prediction of the chaotic time series containing outliers , a hybrid learning method based on the modified fruit fly optimization algorithm ( MFOA) and the least Wilcoxon ( LW) method is proposed in order to train the T-S fuzzy model .The purpose of this is to improve the accuracy and robustness of fuzzy modeling for nonlinear systems .Firstly, the MFOA is used to optimize the antecedent parameters of the Gaussian membership function with the advantages of ease of transformation of such a concept into program code and a high convergence speed , which can improve the identification accuracy and convergence speed in fuzzy modeling .Secondly , the least Wilcoxon method is applied to identify the consequential parameters of the model .When the outliers occur in the training data , the strong robustness of the LW with the outliers is effective for improving the sensitivity of the tradi-tional least mean square method .Finally, a simulation experiment is conducted on the prediction of the Mackey-Glass chaotic time series , and the comparisons of the prediction results by different optimization methods are done to verify the superiority of the modified fruit fly optimization algorithm and in the case of outliers existing , the simu-lation results show the effectiveness and robustness of this proposed method .关键词
修正型果蝇优化算法/最小Wilcoxon方法/例外点/Mackey-Glass混沌时间序列/T-S模糊模型/模糊预测Key words
modified fruit fly optimization algorithm/least Wilcoxon method/outliers/Mackey-Glass chaotic time series/T-S fuzzy model/fuzzy prediction分类
信息技术与安全科学引用本文复制引用
刘福才,窦金梅,王树恩..基于 MFOA 和 LW 的混沌时间序列鲁棒模糊预测[J].智能系统学报,2014,(4):425-431,7.基金项目
河北省自然科学基金资助项目（ F2010001320）． （）
