离散网络系统基于模态的l_2-l_∞滤波器设计

On mode-dependent l_2-l_∞ filtering for network-based discrete-time systems

研究了一种基于网络控制的离散时间系统的能量峰值滤波问题,采用参数不完全已知的马尔科夫链描述由网络引发的系统延时。这样的假设使得滤波器更加具有普遍性,因为现有的方法一般将其延时模态间切换的概率当作完全已知或完全未知的参数来处理。通过把未知的概率参数处理为一个凸空间,可以得到一个基于模态的全阶滤波器。文中利用一个新的实有界引理,得到了可以使相应的误差滤波系统在给定l2-l∞的指标下渐进稳定的LMI充分条件。文中最后给出一个数值算例来展示本方法的有效性。

This paper presents the energy-to-peak filtering problem for a class of network-based discrete- time systems. The network-induced delays are modeled as a Markov chain with partly unknown transition probabilities. Under this assumption, the filter is more general in comparison with the existing methods that take the transition probabilities completely known or completely unknown. A mode-dependent full- order filter is constructed by taking the unknown probabilities into a convex polytope. By using a new bounded real lemma, sufficient LMI conditions are derived to ensure the corresponding filtering errors system is stochastically stable with a guaranteed 12 -1∞ performance index. A numerical example is presented to illustrate the effectiveness of the proposed approach.