Multiagent architecture for online diagnosis of discrete event systems using causal temporal signature approach

Abstract

Industrial environments are becoming more and more dynamic, complex, and uncertain. In such conditions, building safe systems is a major requirement and a big challenge which cannot be achieved without suitable diagnosis tools and methods. The early diagnosis approaches were mostly centralized and done offline. It was then proved that distribution is of a great contribution to realize better diagnosis of complex industrial systems. This paper focuses on Discrete Event Systems (DES) and introduces a distributed approach for online faults diagnosis in DES using Causal Temporal Signatures (CTS). The proposed solution relies on a multiagent architecture where multiple autonomous agents cooperate and coordinate their actions in the diagnosis process.