一种基于T-S模型的快速自适应建模方法

Fast adaptive modeling based on T-S model

在分析 Takagi- Sugeno模型的网格法 (即 ANFIS方法 )、聚类法和模糊树法的基础上 ,提出一种新的改进的建模方法。它划分的子空间的个数与分布和采样数据的特征密切相关 ,具有自适应能力。应用国际标准例题进行仿真 。

Based on analyzing and comparing several Takagi Sugeno fuzzy system modeling methods, a new method of fuzzy modeling based on T S model is presented. Fuzzy subspaces produced have proper numberanddistributionaccording to the sample data distribution and linearizable degree.Parameter identification has proper initial value, so the training process is fast. This new method is applied to a benchmark example, Mackey Glass time series prediction and simulation results demonstrate the effectiveness of the proposed modeling method. It is of less dependence on data and initial values, less calculation amount and higher accuracy and quickly declining error.