Real-Time Monitoring and Early Warning Method of High-Voltage Cable Operation Status Based on Multi-Sensing Fusion and Deep Learning
The traditional real-time monitoring and warning method for the operation status of high-voltage cables can only extract low-frequency signals during abnormal cable operation,resulting in long warning time for fault location.Therefore,a real-time monitoring and warning method for the operation status of high-voltage cables based on multi-sensing fusion and deep learning was designed.A high-voltage cable structural parameter model was established,which includes three parts:the distributed capacitance value of cable conductors and shielding layers,the grounding capacitance current of high-voltage cables,and the grounding induced current of metal shielding layers.Use deep learning algorithms to update the sequence and obtain the updated values of the observed states at the t.After repeated recursion,preprocessed high-voltage cable operation data is obtained.When a single core cable is grounded,its metal sheath will generate induced potential,which will calculate the induced circulating current of the high-voltage cable grounding.The results show that the designed real-time monitoring and early warning method for high-voltage cable operation status based on multi-sensing fusion and deep learning has a positioning error of no more than 2.5 mm for fault location and a warning time of less than 1 s,proving that the proposed method has higher monitoring accuracy.
NETL
NSTL
万方数据
