Modeling and robust diagnosis of power system protection failures using observations in discrete event system
S Reshmila, Rajagopalan Devanathan · 2016
Cascading outages in power system is a major concern in the industry and protection system failures contribute to it. With the increasing complexity of power system, manual monitoring and diagnosis is impossible. Failure diagnosis using discrete event systems has been an active area of interest in the last three decades and has produced creditable results. In this paper, power system protection system is modeled for failure diagnosis using a modified form of the discrete event system model. In particular, a framework using observations in modeling the component is developed for power system protection failure diagnosis. This modeling approach does away with the need to include unobservable events and is capable of predicting the diagnosability during component modeling itself. Simple conditions for diagnosability are stated for the modified DES. A diagnoser is developed for a power system protection example using the modified DES. Further, the robustness of the diagnoser is demonstrated.