论文部分内容阅读
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.
The fuzzy tree method uses the least-squares method to study the consequent parameters of the fuzzy rules and is sensitive to the exception points. For this reason, a minimal Wilcoxon learning method which is insensitive to the exception points is used instead of the least squares method. This method not only improves the shortcomings of the fuzzy tree method sensitive to the exception points, but also inherits the advantages of the fuzzy tree method. The simulation results show that the proposed method can predict the chaotic time series of Mackey-Glass chaotic time series Accurate prediction, verify the validity of the method and the robustness of the exception point.