SDAE-BP Sag Type Identification Method Based on Improved PSO Optimization
In this paper,an optimization method of SDAE-BP sag type recognition based on improved PSO(particle swarm optimization)is proposed.The stackable noise reduction autoencoder(SDAE)in deep learning is used to extract voltage sag signal features,and BP neural network is used to classify the sag data,which solves the problem of unknown features and noise during artificial feature extraction.The improved PSO algorithm was used to optimize SDAE-BP net-work parameters,and the optimization parameters were selected adaptively to improve the network generalization ability and the accuracy of sag type recognition.
万方数据
维普
