Modeling and Simulation of Transformer Under DC Bias Based on Jiles-Atherton Model

A Jiles-Atherton model reflected the characteristic of transformer core hysteresis is established in this thesis, in which difficult parameters identification is the drawback of the model. For identification process is very sensitive to initial values, the relationships between parameters in Jiles-Atherton model and the most representative characters of hysteresis loop are summarized firstly. Hence, the reasonable initial values can be gathered more easily. Then, it is proposed that a parameters identification method based on a sequential algorithm, in which each parameter was optimized in sequence. The mean square error (mse) between experimental data and calculated data affects by each parameter, which is optimization objective should minimize. Compared with existing algorithm, this algorithm is simple, practical and accurate. On the basic, the transformer DC bias model based on the Jiles-Atherton theory is established on MATLAB simulation platform. The simulation results of the hysteresis loops and excitation current are found to be in good agreement with the experimental ones, which strongly proves the effectiveness of this procedure.