Insulator discharge gap distance optimization method based on multi-scale modeling

Abstract

The invention discloses an insulator discharge gap distance optimization method based on multi-scale modeling, and relates to the field of overhead ground wires, and the method comprises the following steps: S1, carrying out data collection, determining parameters, and carrying out electromagnetic simulation calculation through COMSOL; s2, constructing a finite element model according to a calculation result; s3, obtaining a target variable based on the finite element model, processing the sample data, and dividing the sample data into a training sample and a test sample; s4, initializing the multi-scale convolutional neural network; s5, inputting a training sample, and performing model training based on a multi-scale convolutional neural network; s6, after model training is completed, the training effect is verified through the test sample, and the gap distance is determined; the method integrates the capability of a finite element to accurately simulate a physical field and the advantages of multi-scale feature capture and deep data mining of a multi-scale convolutional neural network, achieves the efficient optimization design of the insulator discharge gap distance, and guarantees the safe and stable operation of a power system.

Images