存在多丢包率和传感器多模态故障的网络控制系统可靠H_∞滤波

Reliable H_∞ Filtering for Networked Control Systems With Multiple Packet Dropout Rates and Multiple Sensor Fault Modes

研究了存在多丢包率和传感器多模态故障的网络控制系统的可靠H∞滤波器设计问题.利用服从Bernoulli分布的随机变量描述数据包传输过程中发生的随机丢包,丢包率的数学期望随网络带宽和网络流量的变化在不同取值中动态切换.同时,利用一组在不同故障模态下服从不同随机分布的随机变量描述各个传感器的故障.将滤波误差系统建模成随机切换系统,利用随机系统理论和模式依赖平均驻留时间方法得到了可靠H∞滤波器的设计方法.最后,数值仿真验证了所提方法的有效性.

This paper deals with the reliable H∞ filter design problem for networked control systems with multiple packet dropout rates and multiple sensor fault modes. The random packet dropout phenomenon is represented by a stochastic variable that satisfies Bernoulli random binary distribution, and the expectation changes with the network bandwidth and the network traffic. Meanwhile, the sensor faults are established by using a series of stochastic variables that obey different random distributions at the different fault modes. Then, the filter error system is modeled as the stochastic switch system ; the design method of the H∞ reliable filter is given via the stochastic system theory and the mode-dependent average dwell time method. Finally, a simulation example is used to show the effectiveness of the proposed method.