Abnormal Condition Diagnosis Model of Switchgear Wall Bushing based on Finite Element Simulation and Convolutional Neural Network Algorithm

• Published in: 2023 3rd International Conference on Electrical Engineering and Mechatronics Technology (ICEEMT) pp. 684 - 689
• Main Authors: Liu, Lize, Lin, Dong, Feng, Nan, Li, Guoqiang, Zhang, Xuanyu, Liang, Huishi, Zhang, Ming, Gu, Zhefei
• Format: Conference Proceeding
• Language: English
• Published: IEEE 21.07.2023
• Subjects: Analytical models; convolutional neural networks; fault diagnosis; finite element simulation; Insulators; Software; Software algorithms; Switches; Switchgear; Temperature distribution; temperature field distribution
• DOI: 10.1109/ICEEMT59522.2023.10262901

In view of the phenomenon of local insulation aging and partial discharge in the busbar wall bushing of switchgear, the state diagnosis model of switchgear wall penetration bushings are constructed. Firstly, the electrical-thermal model of the switchgear through-wall bushing is constructed based on the current field theory, and the temperature field is solved by indirect coupling. Then, the precise model of high voltage switch cabinet wall bushing is established based on finite element analysis software, and the simulation analysis is carried out for different types of defects in the bushing to obtain the temperature field distribution of the wall bushing under normal working conditions and different types of defects. Finally, the convolutional neural network (CNN) algorithm was trained based on fault condition data to classify and judge the heat map images and realize the fault diagnosis of the switch cabinet wall bushing.