具有随机长时延与丢包的网络控制系统的随机稳定与镇定

Stochastic Stability and Stabilization of Networked Control Systems with Uncertain Data Losses and Long Delay

对于带有随机性丢包和长时延的网络控制系统,设随机丢包过程为服从有上限的马尔可夫随机过程,随后在离散网络控制系统模型的基础上将随机丢包的网络控制系统建模为带有马尔可夫跳变参数的离散网络控制系统模型。之后将具有随机性丢包和有上限的随机长时延建模为带有长时延的马尔可夫跳变系统。在已有的关于时延马尔可夫跳变系统的基础上,运用Lyapunov函数和LMI进行稳定性分析并提出随机稳定的条件,用锥补线性化方法设计了镇定控制器,最后仿真算例验证了其有效性。

The modeling and stochastic controller design problems of NCS with uncertain Mark- ovian packet-loss and long delay are discussed. The NCS is modeled as a Markovian jump linear system （MJLS） by discrete system method by assuming that the NCS with long delay is subject to a maximum loss of Markov random process on the basis of random packet loss. Moreover, packet-loss dependent stabilizing controller is designed with matrix inequalities （LMIs） and the cone complementarily linearization （CCL） formulation. Finally, an example for a comparison be- tween the two methods is given to demonstrate the effectiveness of the proposed method.