Fault Detection of Networked Control Systems with Packet Dropout
Y.Q. Wang,S.X. Ding,P. Zhang,W. Li,H. Ye,G.Z. Wang
DOI: https://doi.org/10.3182/20080706-5-kr-1001.01501
2008-01-01
IFAC Proceedings Volumes
Abstract:Fault detection of networked control systems with packet dropouts in both sensor-to-controller link and controller-to-actuator link is discussed in this paper. Two types of packet dropouts are considered. One is packet dropout characterized by a Bernoulli process, the other is characterized by a Markov chain. According to the system configuration, packet dropout in the sensor-to-controller link is known to the fault detection system, therefore a time varying parity space based residual generator which is fully decoupled from this influence is designed. Robustness to packet dropout in the controller-to-actuator link which is unknown is achieved by an adaptive threshold. Due to the known probability properties of the residual signal in parity space approach, upper bounds of false alarm rates of the designed thresholds are also given, which are very difficult to obtain in existent observer based fault detection approaches of networked control systems with packet dropout. It is verified that the proposed fault detection approach can also be realized by diagnostic observers by employing the relationship between parity vectors and observers.