New Approach of Diagnosis by Timed Automata

Abstract

The Discrete Event Systems (DES) is a nonlinear dynamic system with discrete state and event evolution. In this article, we are interested in the diagnosis of failures with Timed Automata. The proposed approach is based on the operating time and it is applicable to any system whose dynamic evolution depends not only on the order of discrete events but also on their periods as in industrial processes. The most important part of this work is the construction of a diagnoser which uses observables events to detect and locate the faults. We present on the last part of this work results of the study of the performance of the diagnosis showing the power of this diagnostic approach. An implementation on a hydraulic system is made to illustrate the proposed steps. It put in evidence the effectiveness of this approach. The model of the simulation phase is done using Matlab / Simulink / stateflow.