A methodology for building a fault diagnoser for hybrid systems

Abstract

In this paper, a design methodology for building diagnosers for hybrid systems is proposed. The design methodology uses as a starting point a hybrid automaton model to represent the hybrid system behaviour by means of the interaction of continuous dynamics and discrete events. Then, a hybrid fault diagnoser is designed using the methodology described in this paper and implemented by means of a discrete event system which carries out the mode recognition and diagnostic tasks, both based on residuals generated using models. Both tasks interact each other since the diagnosis module adapts according to the current mode of the hybrid system. The mode recognition task involves detecting and identifying the mode change by determining the set of residuals that are consistent with the current mode of the hybrid system. On the other hand, the diagnostic task involves detecting and isolating faults by identifying the fault that can explain the set of residuals that are inconsistent. A section of the Barcelona sewer network is used as application case study to illustrate the proposed fault diagnosis for hybrid systems.