APPLICATION OF NEURAL NETWORKS IN FAULT CURRENT DETECTION

摘要

In order to effectively achieve fault current detection for a hybrid circuit breaker, design a short-circuit current detection methods based on dynamic neural network, it applies the dynamic neural network to the fault current detection, uses historical memory effect of feedback neural networks to predict and compare the signals, can realize an effective short-circuit current fault detection. Using the Matlab neural network toolbox for simulation, simulation training samples generated by simulation, superpose multiphase harmonic components by single-phase fundamental frequency, similar sine function can be used instead when predigestion, the simulation results show that the method is effective and fast.