Adaptive harmonic separation and detection algorithm based on deep neural network

Abstract

In recent years, with the continuous development of new energy technology, the integration of various new energy power and the popularization of more and more sophisticated electrical instruments, the detection and management of harmonics in the power system has become an urgent problem to be solved. There are a large number of semiconductor materials and asymmetric loads in the system, and the traditional harmonic detection methods are getting worse and worse in the harmonic detection of hybrid new energy power systems. Although the ensemble empirical mode decomposition method has a certain separation effect on harmonics, the spatial and temporal distribution of harmonics in the hybrid power system is quite different, and the artificially set decomposition parameters cannot obtain the optimal decomposition results. In this paper, combined with deep neural network, we propose a harmonic separation detection method that realizes adaptive ensemble empirical mode decomposition by using deep neural network to train an adaptive model. Experiments show that this method can effectively improve the adaptability of the detection system, and can further simplify the detection process and reduce the detection time on the premise of effectively and reasonably separating and detecting each harmonic.