Optimal Design of Double-Fracture Disconnect Switchgears Based on BP Neural Network and NSGA-II Algorithm

GIS double-fracture disconnect switchgears have high requirements on insulation and stress resistance due to their miniaturized and compact design. How to improve the insulation capacity and stress resistance by optimizing the design of key dimensional parameters within a limited volume is the key point of optimization research. In this paper, the finite element method is used to build simulation models of the electrostatic and stress fields of a double-fracture disconnect switchgear. The key dimensional parameters that have a significant impact on the static electric field and stress field are initially analyzed, and these parameters are parametrically processed. In order to solve the optimization problem of large-scale field models using optimization algorithms, the Box-Behnken experimental design method is used to collect samples and establish a BP neural network model. The global Pareto-optimal solution set under multiple objectives is found by combining the BP neural network model and NSGA-II algorithm, and the global optimal solution is obtained according to different objective weights. Under the optimized dimensional parameters, the maximum electric field strength and the maximum principal stress in critical areas are significantly reduced, achieving an effective improvement in the performance of the double-fracture disconnect switchgear.