Research on the Application of LSTM Model Based on Electrical Equipment in Electromagnetic Field Calculation

Abstract

Since the 21st century, the development of information technology and artificial intelligence has reached new heights, and the intelligent manufacture of electrical equipment has become the main development goal of industrial manufacturing in China. The core part of intelligent manufacturing of electrical equipment is the use of data to build mathematical models for real-time analysis, but for large volume and complex structure of electrical equipment, the traditional finite element analysis method will involve large-scale numerical calculations, resulting in slow calculation speed, accuracy is difficult to meet the design requirements, cannot meet the real-time requirements. This paper applies deep learning theory to the electromagnetic field analysis of electric motors, using the electromagnetic field distribution data corresponding to different motor structures to train the built deep learning model to predict the electromagnetic field distribution instead of the traditional finite element calculation, improving research efficiency, reducing time costs and meeting the requirements of intelligent manufacturing for real-time. In this paper, the predicted electromagnetic field distribution is compared with the MSE and MAE of the training set and test set respectively, then the electromagnetic field distribution data is plotted as a line graph to observe the error between the true and predicted values, and finally the predicted electromagnetic field distribution data is visualised as a magnetic line graph using simulation software, and compared with the magnetic line graph obtained from ANSYS Maxwell simulation, and the data of some random points are taken for comparison.