一类非线性系统的神经网络自适应控制

针对一类单输入单输出不确定非仿射型非线性系统,基于多层神经网络提出了一种直接自适应控制方法。该设计方法首先应用多层神经网络自适应模拟逼近逆解中的未知部分,然后应用逆设计和自适应反演设计出虚拟控制量,最后应用反馈线性化设计方法和神经网络设计了直接自适应控制律。并利用Lyapunov稳定性定理推导了神经网络的参数调节律,保证了闭环系统的所有信号均最终一致有界。

Neural adaptive control for a class of nonlinear systems

A direct adaptive control design method is proposed firstly for a class of uncertain single-input single-output (SISO) nonaffine nonlinear system using multilayer neural networks (MNNs). The proposed approach uses MNNs to approximate and adaptively cancel the unknown part of the inverse functions. Then, Inverse design, backstepping design, and feedback linearization techniques are incorporated to design the direct adaptive control law. The adaptive tuning rules for updating the parameters of MNNs are derived by the Lyapunov stability theorem. The developed control scheme guarantees the uniform ultimate boundedness of the closed-loop adaptive system.