To reduce the occurrence rate of accidents during the operation of the heating furnace,a study on electrical fault identification of walking beam heating furnace based on multimodal fusion is proposed.Firstly,a derivative spectrum enhancement model based on nonlinear diffusion is used for signal enhancement processing.Then,perform EMD decomposition on the enhanced electrical signal to obtain all IMF components in the signal.Extract the detailed feature information of electrical faults using wavelet packet transform,and perform multimodal fusion of the detailed feature information of electrical faults.Finally,the Random forest algorithm is used to identify the fault of multimodal fu-sion features.The experimental results show that the loss function value of the proposed method decreases rapidly and can be reduced to below 0.05.The feature extraction ability is strong and the fault identification effect is good,which can help to improve the electrical fault identifica-tion of the walking beam furnace.
维普
万方数据