The Harmonic Currents Detecting Algorithm Based on Adaptive Neural Network

In order to decrease the harmonics and improve the power factors in power system, a detecting algorithm of harmonics and reactive currents based on neural networks and adaptive noise canceling technology is proposed. The structure of neural network and the adaptive weights adjusting algorithm are presented. The contradiction of the detecting speed and the precision has been settled preferably. The proposed algorithm is simulated for detecting the harmonics and the reactive currents of active power filters, Simulation results show that both the tracking speed and steady state error have good effects. The base wave current can be detected in half a period and the steady state error is better. It is benefit to the detecting of harmonics and reactive currents of active power filters in power system.