PQ Disturbances Identification based on Phase-shift and LS Weighted Fusion Combining Neural Network

A new method based on phase-shift and least square (LS) weighted fusion combining neural network was presented for PQ disturbances detection and identification. Through phase-shift and some algebra operations, the method detected the PQ disturbances effectively. By a data dealing process with the detecting outputs, features were extracted for classification. Then five child BP ANNs with different structure were adopted to identify the PQ disturbances. The combining neural network fused the identification results of these child ANNs with LS weighted fusion algorithm finally. Comparing with single neural network, the combining one was more reliable in identification. The simulation results proved the conclusion