基于表面阻抗边界的变压器杂散损耗计算方法

来源 :仪器仪表学报 | 被引量 : 0次 | 上传用户:zzuli666
下载到本地 , 更方便阅读
声明 : 本文档内容版权归属内容提供方 , 如果您对本文有版权争议 , 可与客服联系进行内容授权或下架
论文部分内容阅读
表面阻抗法是基于电磁波理论并将导体表面阻抗作为一种边界条件进行有限元求解的方法, 它具有求解规模小、计算效率高等优势.首先借助工程电磁场数值计算软件Magnet, 使用有限元法和表面阻抗法对该模型进行仿真, 然后基于TEAM Problem 21基准族中的P21-B模型搭建了变压器杂散损耗测试系统, 用于测量在不同频率电流激励条件下的变压器杂散损耗.针对现有测量方法的不足, 引入温度系数对测量结果进行修正, 并采用改进线圈损耗法以得到更准确的杂散损耗测量结果.通过与实验值对比, 验证了有限元法与表面阻抗法计算结果的准确性.从计算规模和时间角度分析, 表面阻抗法远远小于有限元法, 大大节省了计算资源.“,”Based on the theory of electromagnetic waves, the surface impedance method is a finite element method used surface impedance as a boundary condition, whose advantages are of low calculation and high computational efficiency. Firstly, Magnet is used to simulate the model by finite element method and surface impedance. Then, based on the P21-B model in the TEAM Problem 21 reference family, a stray loss test system is built. The stray loss is measured under different current excitation conditions. The temperature coefficient is introduced to improve the measurement, and the improved coil loss method is adopted to obtain more accurate stray loss measurement results. The calculation results of the finite element method and the surface impedance are verified with experimental measurements. The surface impedance is much smaller than the finite element method in terms of calculation scale and time, which greatly saves computational resources.
其他文献
为了提高对辅机故障的事前预知能力,结合深度学习中非监督学习方法的优势,提出基于改进堆叠自编码网络的电站辅机故障预警方法.以辅机的历史正常数据为训练集,利用堆叠自编码
为提高嵌入式时栅角位移传感器测量精度,从传感信号形成机理出发,对短周期误差成因进行了详细分析。通过对绕组等效分析和激励信号分析,确定了短周期误差的主要特性为一次和二次误差,一次误差来源为零点残余误差和直流分量误差,二次误差来源为激励信号正交误差。针对短周期误差补偿,提出了基于超限学习机的误差补偿方法,通过对测量值与真实值样本的训练得到模型最优参数,根据模型参数建立短周期误差模型,利用所得误差模型实