Diagnosis of power system failures using observer based discrete event system

Abstract

With the power system becoming more and more complex it is nearly impossible to monitor and diagnose the failures manually. In this paper power system failure diagnosis is done using the discrete event system approach. Model based diagnosis in discrete event systems (DES) has gone through many developments. We introduce a new framework of Observer based Discrete Event Systems (ODES) to include event outputs in addition to state outputs. It is shown that ODES can make a system diagnosable even though the system is not diagnosable when considering only state outputs and using the conventional DES approach. Besides, ODES while suppressing the unobservable failure states can reduce the severity of state explosion problem inherent in the DES application to large systems.