New Research on Harmonic Detection Based on Neural Network for Power System

Analysis and control for power quality by neural network is a new research field in electrical power system. Rapid and reliable extract the harmonic components determine the overall performance of Active Power Filter (APF). This paper presents a new three-layer feedforward neural network based on error back-propagation algorithm that the training sample without time delay, which can detecting harmonics for power system in real-time. With the simulation study using Matlab, the simulation results illustrate that the harmonic detection method based on neural network is feasible, which can quickly detecting the harmonics for non-linear load.