不确定离散线性系统鲁棒ESPR估计新方法

Robust ESPR estimation for uncertain discrete-time linear systems--an improved LMI approach

在时域中考虑多胞不确定离散系统的鲁棒ESPR(Extendedstrictlypositivereal)估计问题.基于最近的结果首先提出了离散线性系统具有ESPR性能的新判据,该判据实现了系统矩阵和Lyapunov矩阵的分离,这样在系统分析和综合时可以采用参数依赖的Lyapunov函数.基于提出的新判据和LMI(Linearmatrixinequality)技术给出了多胞不确定系统具有ESPR性能估计器的设计方法.与传统的基于二次稳定方案相比,提出的设计方法具有较小的保守性.

The problem of extended strictly positive real (ESPR) estimation is considered in time domain. Based on recent results, a new criterion for ESPR performance is proposed for discrete-time linear systems. The proposed criterion exhibits a kind of decoupling between the system matrices and the Lyapunov matrix, so the use of parameter-dependent Lyapunov functions is allowed for analysis and synthesis problems. Based on the new criterion and LMI technique, robust estimators with ESPR are designed for polytopic uncertain discrete-time systems, with the result showing that less conservativeness is achieved compared with the results based on the quadratic framework.