Moving Average control chart for the detection and isolation of temporal faults in stochastic Petri nets

Abstract

This paper deals with problems of detection and isolation of temporal faults in timed stochastic discrete event systems. Partially labeled timed Petri nets are used to model the considered systems. Temporal faults corresponding to significant variations of the support of the probability density function (pdf) are considered. A pdf represents the firing duration of each transition. A Moving Average control chart (also known as a Moving Mean chart) is applied in order to detect the variation of mean duration. The advantages of the proposed analysis are to detect variations in time series when parameters vary slowly and to isolate the faults thanks to the signature table.