Research on Intelligent Diagnosis Method for High Voltage Cable Insulation Faults Based on Multi-Sensory Fusion

Aiming to address the problems of high detection cost, poor adaptability, and low accuracy of the existing high-voltage cable insulation fault diagnostic system, an intelligent diagnostic method for high-voltage cable insulation faults based on multisensory fusion is proposed. First, the detection method effectively decomposes the acquired data information through variable modal decomposition (VMD) and obtains relevant feature quantities; secondly, the current cable insulation fault is predicted according to the feature quantities using the long-short memory neural network (LSTM) algorithm. The experimental results show that the method has good prediction performance, and the detection accuracy can reach 97.02%, which is significant for guaranteeing safe power grid operation.