一种面向非线性系统控制的自适应模糊小脑神经网络及研究

Research on Adaptive Non-linear Control Method Based on Fuzzy CMAC

利用Lyapunov自稳定性准则,将自适应机制引入到模糊小脑神经网络(CMAC)的实时学习算法之中,提高其在闭环控制系统中的鲁棒性,使其能够有效地对模型未知的非线性系统进行实时控制.仿真表明自适应CMAC神经网络由于采用了基于Lyapunov自稳定准则的学习算法,系统的跟踪稳定性和误差收敛性都能够得到保证,而且不需离线学习阶段,实时控制效果较好.

Adaptive mechanism is introduced into fuzzy CMAC network based on Lyapunov stability equation to enhance the robustness of model-unknown nonlinear system controller performance. The simulation results reveal that the fuzzy CMAC which learning method based on the Lyapunov stability equation is capable to guarantee the system trace stability and error convergence, don＇t need off-line learning phase and makes good performace.