Harmonic and Interharmonic Detection and Analysis of Neural Network Power System Based on Big Data

Abstract

Aiming at the low measurement accuracy and poor time lag of traditional power system harmonic and interharmonic detection methods, a new power harmonic detection method is proposed, namely, the neural network power system harmonic detection analysis based on big data. It combines wavelet transform with artificial neural network, uses wavelet transform to extract the characteristic components in the signal. And then separates and extracts the different frequency band signals, and then uses the corresponding artificial neural network to analyze and identify the extracted frequency band signals to obtain each Harmonic type, amplitude and phase parameters in the hierarchy. The final simulation experiment shows that this research proposes to combine wavelet transform with artificial neural network to give full play to the time and frequency locality advantages of wavelet transform, as well as the advantages of artificial neural network self-adaptability and model maximum value advantages, which are effective for periodic changes and interference in electric power. It has strong tracking performance and recognition ability, which further shows that the power harmonic detection method proposed in this study is feasible and effective.