基于动态结构自适应神经网络的不确定时滞系统跟踪控制

Tracking control for uncertain time-delay systems based on dynamic structure adaptive neural network

针对一类不确定时滞非线性系统,提出一种自适应跟踪控制器.首先采用Lyapunov-Krasovskii函数设计时滞补偿器,并构造其中的参数调节规律.再针对建模误差及不确定非线性,引入动态结构自适应神经网络,其隐层神经元个数可以随着跟踪误差的增大而在线增加,以提高逼近精度.最后,用仿真示例表明本文所提方法是有效的.

For nonlinear systems with uncertain time delays, an adaptive tracking controller is presented. We use the Lyapunov-Krasovski function to design the time-delay compensator and construct its coefficient updating rules. A dynamic structure adaptive neural network is adopted to eliminate the unknown nonlinearities and modeling error. The simulation shows that the presented method is effective.