非正弦激励下高频变压器铁心动态磁滞特性模拟方法

Simulation of Dynamic Hysteresis Characteristics of High Frequency Transformer Core Under Non-sinusoidal Excitation

高频变压器是DC/DC变换器的核心部件,其激励波形一般为非正弦波,工作频率可达数千赫兹以上,导致铁心损耗显著增加,准确模拟非正弦激励下高频变压器铁心动态磁滞特性,有助于铁心损耗计算以及变压器的电磁暂态分析。首先,该文考虑铁心磁化物理过程,利用系统级电力电子仿真软件PLECS中底层模块和语言编写程序,建立基于静态J-A磁滞模型和磁导-电容类比法的铁心静态磁滞磁导模型;然后,考虑高频条件下涡流效应与弛豫效应对铁心动态磁滞特性的影响,根据损耗统计理论建立铁心动态磁滞磁导模型;最后,基于模拟退火与Levinberg-Marquardt混合算法提取静态J-A磁滞模型参数,根据铁心损耗特性曲线采用最小二乘法提取动态损耗系数。对比分析不同占空比、高频方波电压激励下电流波形以及磁滞回线的仿真结果与实验测量结果,验证了模型的有效性。

The high-frequency transformer is the core component of the DC/DC converter.Its excitation waveform is generally non-sinusoidal,and the operating frequency can reach more than thousands of hertz,resulting in a significant increase in core loss.Accurately simulating the dynamic hysteresis characteristics of the high-frequency transformer core under non-sinusoidal excitation is helpful for the calculation of core loss and the electromagnetic transient analysis of the transformer.In this paper,firstly,considering the physical process of core magnetization,the static hysteresis permeability model of core based on static J-A hysteresis model and permeance-capacitance analogy method is established by using the bottom module and language in the system-level power electronic simulation software PLECS.Then,considering the influence of eddy current effect and relaxation effect on the dynamic hysteresis characteristics of the core under high-frequency conditions,the dynamic hysteresis permeance model of the core is established according to the loss statistical theory.Finally,the static J-A hysteresis model parameters are extracted based on the hybrid algorithm of Simulated Annealing and Levinberg-Marquardt,and the dynamic loss coefficient is extracted by the least square method according to the core loss characteristic curve.The simulation results and experimental results of current waveform and hysteresis loop under different duty cycles and high-frequency square wave voltage excitation are compared and analyzed,which verifies the effectiveness of the model.