万慧, 齐晓慧. 线性/非线性切换扩张状态观测器的收敛性和稳定性分析[J]. 信息与控制, 2020, 49(2): 163-169. DOI: 10.13976/j.cnki.xk.2020.9073
引用本文: 万慧, 齐晓慧. 线性/非线性切换扩张状态观测器的收敛性和稳定性分析[J]. 信息与控制, 2020, 49(2): 163-169. DOI: 10.13976/j.cnki.xk.2020.9073
WAN Hui, QI Xiaohui. Convergence and Stability Analysis of Linear/Nonlinear Switching Extended State Observer[J]. INFORMATION AND CONTROL, 2020, 49(2): 163-169. DOI: 10.13976/j.cnki.xk.2020.9073
Citation: WAN Hui, QI Xiaohui. Convergence and Stability Analysis of Linear/Nonlinear Switching Extended State Observer[J]. INFORMATION AND CONTROL, 2020, 49(2): 163-169. DOI: 10.13976/j.cnki.xk.2020.9073

线性/非线性切换扩张状态观测器的收敛性和稳定性分析

Convergence and Stability Analysis of Linear/Nonlinear Switching Extended State Observer

  • 摘要: 线性/非线性切换扩张状态观测器(linear/nonlinear switching extended state observer,SESO)是为追求更高的控制精度和抗扰能力,在进行状态估计时根据状态误差大小在线性扩张状态观测器(linear extended state observer,LESO)和非线性扩张状态观测器(nonlinear extended state observer,NLESO)之间进行切换的一种扩张状态观测器.但由于该算法存在非线性结构,目前还没有从频域角度对该方法进行收敛性和控制系统稳定性分析的相关研究.针对此问题,以二阶系统为研究对象,首先对SESO的结构及参数选择进行了介绍,然后,分别采用等效增益法和描述函数法,从频域角度对了其状态估计误差的收敛性和基于SESO的自抗扰控制系统的稳定性进行了研究.结果表明,三阶SESO具有良好的收敛性,能实现对系统状态和"总扰动"的误差估计,并且选择适当的SESO参数,可避免极限环的产生.

     

    Abstract: Linear/nonlinear switching extended state observer (SESO) is a newly proposed state error-based switching observer that switches between the linear extended state observer (LESO) and the nonlinear extended state observer (NLESO) when performing state estimation in order to realize high control precision and strong anti-disturbance ability. Presently, there are no studies on the convergence and stability analysis of this observer in the frequency domain. Considering this issue, the structure and parameter tuning of SESO is introduced for a second-order plant; then the convergence and stability of the observer are analyzed in the frequency domain using the equivalent gain method, and the function is described. The results show that the third-order SESO has good convergence and can estimate the states and "total disturbance" of the system, and that if the parameters are tuned properly, limit cycles will not exist.

     

/

返回文章
返回