On designing a better operator assistant for emergency control in power systems

Why don't power system operators use current emergency control algorithms when dealing with emergencies? The algorithms should assist the operator but what they do differs from what a human assistant would do. This paper develops the desirable characteristics of an ideal assistant from the human model, and uses these characteristics to design an emergency control module for a power system operator assistant (OA). The most desirable characteristic of an assistant-understanding and predicting the needs and actions of the operator-leads to the development of an operator-centered system state classification and a generic OA structure. The emergency control module of the OA employs intelligent methods like embedded heuristics, rules and fuzzy logic. The key features of the methodology are: solutions the operator can understand and implement intelligent problem characterization, selection among several heterogenous control action tools (CATs), and computation time constrained by the nature of emergency.