Remote fault detection system design with online channel reliability information

Abstract

In this article, the remote fault detection (FD) problem over unreliable communication channels encountering bit errors and quantisation errors is studied, where data transmissions can be centralised or decentralised. The online channel reliability information represented by the so-called bit error rate (BER) is first analysed in the view of control engineering and then transformed into time-varying parameters and stochastic signals in the control systems. By using the information of BERs provided by communication systems, we propose a new remote FD system which consists of an observer-based residual generator and a residual evaluator. The design of the residual generator is formulated as a model matching problem such that an optimal FD performance can be achieved. Then the residual signals are evaluated and compared with a threshold to detect the occurrences of faults. The false alarm rate is guaranteed to be below a given level. All the solutions are presented in terms of linear matrix inequalities. A numerical example is used to illustrate the effectiveness of the proposed methods.

Keywords:

• fault detection
• communication systems
• threshold selection
• false alarms
• networked control systems

Acknowledgement

The work was supported by the German Research Foundation (DFG) under Grant DI 773/10.