论文部分内容阅读
在传统T-S模型的基础上,提出一种扩展T-S模型.该模型由一组模糊规则组成,由规则前件实现输入空间的划分,将成员函数及其函数变换引入规则后件以实现对输入予空间的非线性映射.对于该模型的建立,使用改进量子遗传算法优化规则前件,递推最小二乘法确定规则后件参数.通过对两个典型非线性系统辨识,仿真结果表明了该模型可以显著提高辨识精度,且具有很好的泛化性能.
Based on the traditional TS model, an extended TS model is proposed, which is composed of a set of fuzzy rules. The former is used to divide the input space and the other is to introduce the member functions and their functions into the rules. Space for the establishment of the model, the use of improved quantum genetic algorithm to optimize the rules of the former, recursive least squares method to determine the parameters of the rules after the two typical nonlinear system identification, the simulation results show that the model can Significantly improve the identification accuracy, and has good generalization performance.