An On-site Fault Remote Intelligent Diagnosis and Warning Method for Electricity Information Collection Terminal Based on Deep Learning

An on-site fault remote intelligent diagnosis and warning method for electricity information collection terminal based on deep learning is proposed to address the complex forms of faults and the difficulty of manual inspection in meeting operation and maintenance needs. Firstly, model the fault diagnosis and warning system for the electricity information collection terminal, including the state monitoring layer, prediction and diagnosis layer. Then, the deep forest algorithm is improved by setting weights and applied to remote real-time intelligent diagnosis of faults in electricity information collection terminals. Finally, a BLSTM-GRU model is constructed by combining bidirectional long short-term memory network (BLSTM) and gated recurrent unit (GRU) networks, and applied to fault prediction of electricity information collection terminals, as well as issuing warnings based on the prediction results. Based on the selected data samples, experimental analysis is conducted on the proposed method, and the results shows that its fault diagnosis accuracy reaches 96.03%, and the fault warning results are reliable.