Observer-based fault detection for networked control systems with network Quality of Services

This paper addresses the problem of the fault detection for linear time-invariant systems over data networks with limited network Quality of Services (QoS). An integrated index ηkηk, which related with data dropout, network-induced delay and error sequence, is presented to described the non-ideal QoS, the probabilistic switching between different ηkηk is assumed to obey a homogeneous Markovian chain. Then by view of the augmented matrices approach, the fault detection error dynamic systems are transferred to Markov jumping systems (MJSs). With the developed model and using the bounded real lemma (BRL) for MJSs, an H∞H∞ observer-based fault detection filter is established in terms of linear matrix inequalities (LMIs) to guarantee that the error between the residual and the weighted faults is made as small as possible. A simulation example is provided to show the effectiveness of the present methods.