Intermittent failure diagnosis in industrial processes

Abstract

Intermittent failure diagnosis based on discrete event system (DES) in industrial processes is studied. Intermittent failures are defined as failures that can automatically recover once they have occurred. In this paper this sort of failures are modeled and some characteristic parameters are defined. The original approach to diagnose permanent failures is revisited and extended in order to include new diagnosis skills in the diagnosers. The developed methodology is applied to a classical industrial subprocess system composed by a pump and a valve. The developed diagnoser and this example are simulated in Matlab, employing Simulink for the continuous model and stateflow for the DES diagnoser.