Diagnosability of discrete event systems for temporary failures

Failure Detection and Diagnosis (FDD) using Discrete Event System (DES) framework is used for wide range of applications because of simplicity of both the model and associated algorithms. Initial research was focussed on permanent failures. Many systems exhibit temporary failures in the sense that system recovers to normal condition after failure. Contant et al. extended the event based DES framework, developed for permanent failure by Sampath et al. to handle temporary failures. Zad et al. developed a state based DES framework for permanent failures which has several advantages compared to the event based approach. In this paper, we extend the state based DES framework, developed for permanent failures, to handle temporary failures, maintaining the same order of complexity. The proposed DES framework has several advantages compared to that by Contant et al. namely, capability to detect failures that occur before starting execution of the diagnoser, same diagnoser for both temporary and permanent failures, etc.

Graphical abstractFigure optionsDownload full-size imageDownload as PowerPoint slideHighlights► Unified scheme for DES diagnosability of temporary and permanent faults is proposed. ► Complexity of unified scheme is equivalent to the scheme for permanent faults only. ► Working of the scheme is illustrated on two benchmark systems.