基于Dahlin算法的CMAC控制器设计

针对传统的基于Dahlin算法的控制器在对变时滞系统进行控制时控制效果恶化,甚至发生不稳定现象的弱点,提出了以CMAC神经网络与Dahlin算法相结合的控制方法。以CMAC神经网络作为一个前馈控制器,通过对Dahlin控制器输出的学习,实现时滞系统的自适应稳定控制。仿真实验表明,这种复合控制方法保留了Dahlin算法与CMAC神经网络的各自特长,同时具备学习速度快,适应能力强的优点,具有良好的稳定性控制效果。

CMAC Controller Based on Dahlin Algorithm for Time-delay Systems

To the problem that the conventional controller based on Dahlin algorithm may destabilize a control system when the delay time varying,an adaptive controller based on self-tuning principle combining Dahlin algorithm with CMAC is proposed.The CMAC neural network as feed forward is introduced,using the output from Dahlin controller as input parameters.This feature may avoid the influence of time delay changing.Simulation result shows that the extend self-tuning controler works better than the classical Dahlin algorithm.