Diagnóstico de Sistemas de Eventos Discretos Controlados: Un Enfoque Basado en Crónicas y Análisis Modular Usando Modelos de Autómatas

Abstract

Nowadays industrial process systems are becoming more complex and it is needed simpler and efficient diagnosers by decreasing the dimension of their models. Modular diagnosis has proved to be very efficient in order to reduce the complexity asociated to discrete event systems. This work proposes a diagnostic approach based on chronicles and modular temporized analysis. Each fault is associated with a set of chronicles and each chronicle recognizes fault signature which is obtained from the state diagnoser associated to the finite state automata defined for each process module. A base of modular chronicles is created, which are manipulated through a coordination protocol that runs online in order to produce the global diagnosis. The performance of the proposed approach is tested on a industrial case of study and adequate results are reached.