Power Quality Disturbance Classification Method Based on Transit Search Optimization CNN/BI-GRU
Aiming at the problem of low identification accuracy of complex power quality disturbance classification methods,a method for power quality identification and classification based on the transit search optimized multi-modal network model was proposed.Firstly,the Gramian angular field was used to perform data processing on the initial one-dimensional time series signal to obtain two-dimensional image data.Secondly,the time series signal and image data were input into the multi-modal network for feature extraction,and the transit search algorithm was used to optimize the parameters of the multimodal network to improve the feature capture capability of the network.Then,through the feature fusion module,the time series features and image features were fused effectively.Finally,the self-attention mechanism was used to enhance the network model′s ability to understand contextual information.The results showed that the method proposed in this paper had a classification accuracy of 99.2%in a noise-free environment,and an average classification accuracy of 98.3%in different signal-to-noise ratio environments.The proposed method achieves accurate classification of increasingly complex power quality disturbances in new power systems,and it is more robust than traditional classification methods.
power quality disturbancedeep learningGramian angular fieldfeature fusiontransit search algorithmself-attention mechanism
国家科技期刊平台
NSTL
万方数据
