Abductive Reasoning Network-Based Diagnosis System for Fault Section Estimation in Power Systems

This study presents an abductive reasoning network (ARN) for real-time fault section estimation in power systems. The proposed ARN handles complicated and knowledge-embedded relationships between the circuit breaker status (input) and the corresponding candidate fault section (output) using a hierarchical network with several layers of function nodes of simple low-order polynomials. The relay status is then further used to validate the final fault section. Test results confirm that the proposed diagnosis system can obtain rapid and accurate diagnosis results with flexibility and portability for diverse power system fault diagnosis. In addition, the proposed method performs better than the artificial neural networks (ANN) classification method both in developing the diagnosis system and in estimating the practical fault section. Moreover, this study demonstrates the feasibility of applying the proposed method to real power system fault diagnosis.