Feature Extraction Method of Power Quality Disturbances Based on PSO and VMD

The power quality is an important index of power industry. It is important to identify the type of power quality disturbances (PQDs) accurately. To address the low precision of PQDs feature extraction and the complexity of recognition models, a novel method is proposed in this paper, which is based on variational mode decomposition (VMD) optimized by particle swarm optimization (PSO). Firstly, to obtain more independent intrinsic mode functions (IMF), the penalty factor in VMD is optimized by PSO. Secondly, the endpoint effect in extracting time-domain features is mitigated through a mirror extension method. Finally, the optimal feature vector is created by extracting the peak value and power spectrum entropy of different IMFs, which is input into a backpropagation (BP) neural network for classification. The simulation results show that the classification accuracy of the proposed feature extraction method for eleven kinds of PQDs is 97.88%, compared with the other models, the classification accuracy is the higher in different signal to noise ratio (SNR) environments.