A Fault Diagnosis System For Substation Based On Optimizing Combinations Of Fault-Masses

It is important for power system restoration to identify the fault section, whenfault occurs. The task becomes challenging because of some protective devicenon-operations or circuit breaker non-operations and some faults of sensors andcommunication equipment. A new method for fault diagnosis based on optimizing thecombination of fault-masses is presented in this paper. Protective device non-operations and circuit breaker non-operations are thoughtas faults. One group made up of a fault source and some related protective devicenon-operations or circuit breaker non-operations is called fault-mass; and a series ofsymptoms resulted from a fault-mass is called symptom-mass. Each fault-mass iscorresponding with a symptom-mass and a fault black-out field. The entire field ofsubstation is plotted into many fields represented by nodes, which are connected withothers by branches, and all protections are modeled by the protection model. Faultsare set and analyzed based on this substation network, so fault symptoms of the faultsare acquired. The concept of effective fault-mass is presented with the help of the known faultblack-out fields and fault symptoms, and the fault diagnosis problem is transformedinto matching the known fault symptoms with some effective fault-masses. Candidateanswers are evaluated by the fitness function, and the genetic algorithm is adopted tosearch out the optimal answer from the combinations of effective fault-masses. The developed methodology has the ability to deal with complicated faults, andthe diagnosis results include not only fault sources, but also device non-operationsand missing signals. It is beneficial for the operators to confirm broken-downcomponents quickly.