A Neural Algorithm Model Based on Niche Technique and Its Study of Fault Diagnoses

The studies improved the algorithm of back propagation neural network and analyzed the niche technique. Based on the above studies, we proposed a neural algorithm model through the niche genetic technique. After trained and tested in its performance, this algorithm model was applied to the fault diagnoses of power transformers. The experimental data show that the algorithm model can converge quickly, efficiently diagnose the faults of power transformers, and increase greatly ratio of fault recognition. This algorithm model could have some reference value for fault diagnoses of similar electrical equipment