Towards diagnosing cascading outages in cyber physical energy systems using temporal causal models

Abstract

Cascading failures in critical cyber physical systems such aspower systems are rare but lead to huge social and economicimplications. Timely diagnosis of faults in these systems isa challenging task due to inherent heterogeneity and scale ofthe system. In the past, we have successfully demonstrateda robust technique for diagnosing independent componentfaults using Temporal Causal Diagrams (TCD) at sub-systemlevel. In this paper, we present a systematic approach of usingthe sub-system level fault models to auto-generate a systemlevelfault model that helps in diagnosing cascading failures.We show the time complexity of our model generation algorithmusing industry standard Power Transmission networks.Further, we describe the updates to the existing TCD reasoneralgorithms and report the TCD diagnosis results for simulatedmulti fault scenario on a standard power system.