New expert system for fault diagnosis in middle Delta Electricity Zone, Egypt

This paper presents a new expert system for fault diagnosis in power system. For this expert system, expertise is represented by propositional logic and it is converted into one Boolean function. By applying Quine-McCluskey tabular method, all Prim Implicants (PIs) of the function can be obtained automatically. The method estimates faulty sections using current state of protective relays by just checking an intuitive look-up table. The paper proposes the application of this approach for fault diagnosis in 66 kV network of Delta Electricity Zone, Egypt. The development procedure for this new expert system is addressed. Verification of the relay and CD information is performed based on Error Detection and Correction Table (EDCT) using PIs that contain only facts, or state of protective devices. Therefore, off-line inference is possible by off-line identification of PIs, which greatly reduces the on-line inference to fit real-time applications. The results show that the developed logic-based expert system is highly efficient.