An on-line expert system for fault section diagnosis in power systems

This paper presents an expert system for diagnosis of power system fault allocation in real time (SIDUF-TR). The system uses information on the tripped relays and circuit breakers to identify the most probable faulted element of the power system, serving as a decision making support for energy control center dispatchers. First, the expert system structure is presented, including a description of the inference method used to determine the most probable failures places. Then the architecture for real-time operation of SIDUF-TR in a control center is described. Finally, the result of the application of the system to a real disturbance is presented.