基于遗传算法和神经网络的真空灭弧室电场优化方法研究

打开文本图片集

关键词：真空灭弧室；电场；正交实验；神经网络；结构优化

DOI：10.15938/j. emc.2025.09.014

中图分类号：TM561 文献标志码：A 文章编号：1007-449X（2025）09-0147-12

Abstract ：Reasonable arrangement of the structure of the interrupter is an effctive way to avoid insulation damage. A structural optimization method based on neural network and genetic algorithm was proposed to optimize the insulation performance of vacuum interrupter. The electric field model of 12kV vacuum interrupter was built.The electric field distribution of the main shield with different parameters was calculated.The orthogonal regresson method and BP neural network were combined to construct the mapping model of the structural parameters of the main shield and the maximum field strength and field strength uniformity of the vacuum interrupter. The genetic algorithm was introduced to optimize the structure of the main shield. The influence of shielding material on the insulation performance of vacuum interrupter was analyzed. The results show that the electric field of the vacuum interrupter decreases with the increase of the diameter of the shield，and increases with the increase of the thicknessof the shield，and decreases with the increase of the angle. Stainless steel as the main shield material can make the vacuum interrupter have the advantages of low cost and excelent insulation performance. The maximum electric field of vacuum interrupter is reduced by 2.07×106V/m ，and the uniformity of electric field at shield and contact is increased by 1.4×104V/m after optimization. The optimization design method can reasonably improve the structural of vacuum interrupter and efectively control the electric field distribution of vacuum interrupter.

Keywords：vacuum interrupter； electric field;orthogonal experiment； neural network； structural optimi-zation

0引言

随着社会不断进步，我国对电力设备的需求与日俱增。（剩余15610字）