A novel method for power quality comprehensive evaluation based on ANN and subordinate degree

A comprehensive evaluation approach of power quality (PQ) based on subordinate degree-BP neural network was proposed in this paper. In case the BP neural network training process is trapped by the local minimum point, genetic algorithm (GA) was introduced to optimize the network's initial weights. A large number of samples based on the random-distribution theory were produced to train the network, and the network output results were analyzed according to the subordinate degree rule. Compared with the BP network neural method, the proposed subordinate degree-BP neural network method can evaluate the PQ level correctly and analyze all kinds of PQ indices exactly. By practically evaluating the 0.38kV distribution network, the proposed approach is proved correct and feasible.