基于T-S模糊模型的非线性自适应逆控制系统

现在非线性自适应逆控制系统中，由于对象模型和逆控制器均采用非线性自适应滤波器，自适应过程需要同时训练至少两个串联的非线性自适应滤波器，从而造成自适应学习过程过于复杂。利用一组自适应LMS滤波器建立非线性对象的T-S模糊模型，它为逆控制器的学习提供了准确的解析的对象模型Jacobian信息，从而有效简化了自适应逆控制学习过程。仿真结果表明，无论是离线建模还是在线建模，该非线性自适应逆控制方法均能取得理想的控制效果。

Nonlinear Adaptive Inverse Control System Based on T-S Fuzzy Model

A nonlinear adaptive inverse control system(NAICS) needs at least two nonlinear adaptive filters,one of which is used as identifier and the other as a controller.It is difficult to train such two filters simultaneously because the identifier cann't supply the controller precise gradient.In order to simplify the adaptive learning process of NAICS,a NAICS based on T-S fuzzy model was built.In the new NAICS,the identifier was replaced by a T-S fuzzy model,which was built by a set of adaptive LMS filters.Both the structure and the learning algorithm of LMS filters are simpler than those of nonlinear adaptive filters.Particularly a T-S fuzzy model can provide controller precise Jacobian information,so the learning algorithm of the controller is significantly simplified.Simulation results show the new NAICSs has good performance.