Abstract:

The accurate prediction of transformer core hysteresis characteristics and the reliable identification of its model parameters have been a difficult problem for scholars at home and abroad. In view of the existing mainstream magnetization Jile-Atherton(J-A) models, there are many identification parameters, long calculation time, and easy to fall into the local optimal solution. A J-A hysteresis model parameter identification method based on constriction factor improved particle swarm optimization(PSO) algorithm is proposed. The J-A static hysteresis inverse model with magnetic induction intensity as the input variable was established, and a dynamic hysteresis model considering the inclusion of eddy current loss and anomalous loss factors is proposed. An improved PSO optimization algorithm based on the constriction factor is proposed, which can realize the fast identification of key parameters of the J-A hysteresis model, in order to address the problems of low computational accuracy and not easy to find the optimization quickly. The limitation of traditional particle is overcome by the algorithm, which has the characteristics of both global and local optimization, and is easy to realize the fast parameter identification of J-A magnetism model. The reliability of the proposed improved PSO algorithm for different magnetic density peak operating conditions is verified through simulation case analysis, and the iterative convergence speed and accuracy are better than those of the conventional PSO algorithm.

Key words: Jile-Atherton model, parameter identification, particle swarm optimization algorithm, hysteresis loop