Design of a large-scale outage fault mining algorithm based on correlation analysis of instantaneous power features

A large-scale outage fault mining algorithm based on instantaneous power feature correlation analysis is designed. By deleting the vertices corresponding to tripped circuit breakers and their neighboring edges, all such passive regions are output, and the large-scale outage regions are identified by combining the concepts of cut-point and cut-set in graph theory. In view of the complex spatial and temporal characteristics of large-scale outage faults, a passive two-port network model of the outage region is constructed to analyze the current and voltage characteristics of transient and permanent faults, and the gray correlation is used to calculate the correlation between the longitudinal and transversal characteristic sequences, magnitude, and energy of the faulted line. Combined with the magnitude aberration degree and the resolution coefficient, the problem of accurately capturing the voltage and current characteristics of the fault instance is solved by considering both vertical and horizontal directions. The fault degree of components is obtained by normalization, and the accurate mining of outage faults is realized according to the set threshold value by considering the time dimension change and the interrelationship between lines. Through experiments, it is proved that the proposed algorithm can realize large-scale outage fault mining and reduce misjudgment and omission, and the mining error is below 0.10%, which improves the safe operation of the power grid.