Multi-objective structural optimization of UHV composite insulators based on pareto dominance

Abstract

In order to improve the electric field intensity distribution of the composite insulator, it is necessary to reduce the potential gradient at both ends of the surface and the insulator junction. Therefore, the composite insulator structure should be reasonably optimized. At the same time, the grading rings are adopted. The structural parameters of composite insulator and the location and size of the grading ring are selected as optimization targets. Two optimization objectives are the minimum peak value and the uniformity of the electric field intensity along the insulator surface. Considering the complexity and economy of composite insulators with different sizes, a composite insulator optimization model based on rough set theory is established. Furthermore, a multi-objective particle swarm optimization-back propagation artificial neural network (MOPSO-BP) algorithm based on Pareto dominance method is proposed to solve the composite insulator optimization problem. The result of the Pareto solutions is evaluated by the finite element analysis software COMSOL. The result shows that the surface electric field intensity of the optimized insulator and the grading ring are both smaller than the results before optimization.