基于least-squares算法的随机非线性系统自适应控制的模块化设计

The modular approach of adaptive control of stochastic nonlinear systems based on least-squares algorithm

线性系统中基于估计的模块化设计就是任意镇定的控制器能够和任意标准的辨识器结合在一起,将这种设计思想推广到不确定随机非线性系统的鲁棒自适应控制设计中,设计一个输入状态稳定控制器模块,该控制器具有很强的参数鲁棒稳定性.根据Swapping技术设计辨识器模块,它由两个滤波器组成,将动态参数模型转化为静态模型,基于最小二乘(least-squares,LS)算法设计参数自适应律,并与梯度算法的设计结果进行比较研究,仿真结果验证了该算法的有效性.

In the linear systems the estimation-based modular approach is that any stabilizable controller can combine with any standard identifier. The estimation-based modular approach was extended to the robust adaptive control design of uncertain stochastic nonlinear systems. An ISS controller module with strong parametric robust stability was designed. According to the Swapping Technique, an identifier module was designed which consisted of two filters that converted a dynamic parametric model into a static one. Based on least-square algorithm, the parameter update laws were designed and compared with the result of gradient algorithm. The simulation results show the validity of the algorithm.