逆系统方法的径向基函数网络实现

研究采用径向基函数网络（ＲＢＦＮ）构造系统逆控制器的工程实现问题，同时给出该直接逆动态控制器存在的充分条件。为进一步改善基于ＲＢＦＮ的直接逆动态控制器的动态性能，对该伪逆系统进行ＰＩＤ综合。仿真研究表明，以ＲＢＦＮ拟合对象逆过程的ＰＩＤ综合控制策略不仅能改善系统的动态性能，而且具有良好的参数鲁棒性能。

Realization of Inverse Systems Based on Radial Basis Function Neural Network

Realization of inverse controller based on radial basis function neural network and sufficient conditions of such RBFN inverse dynamic controller are investigated. In order to improve the dynamic performance, PID regulator is introduced in pseudo linear system including RBFN inverse controller. With the help of simulations, it can found that the control strategy based on RBFN inverse controller possesses advantages of approximation and identification robustness.