观测器为基础的多包传输网络化系统故障检测(英文)

Observer-based fault detection for networked systems with multiple packets transmission pattern

研究了基于不确定马尔科夫模型的多包传输网络化系统故障检测问题.通过引入传输矩阵形式,部分转移概率矩阵元素未知的马尔科夫跳系统模型被用来描述相邻采样周期不同数据包传输时的相关特性.基于获得的模型,利用李雅普若夫方法和随机分析技巧,建立了线性矩阵不等式形式充分条件,所得扩展的误差系统满足随机稳定性和H∞扰动水平.特别的,本文的结论可以包含经典马尔科夫模型和切换模型作为其特例.最后通过仿真例子说明了所用方法的有效性.

Fault detection problem of multiple packets transmission networked systems with uncertain Markovian jump system model is investigated. By introducing transmission matrix, we build Markovian jump system model with partly unknown transition probabilities to describe the interaction between different packet transmissions in adjacent sampling periods. Based on the obtained model, with the help of Lyapunov method and stochastic analysis techniques, we design mode-dependent fault detection filter in terms of linear matrix inequalities to stabilize stochastically the augmented error system in H-infinity framework. Corresponding results of classical Markovian jump system model and the switched system model can be included as our special cases. The effectiveness of the proposed method is demonstrated by simulation examples.