电工技术学报  2023, Vol. 38 Issue (15): 4063-4075    DOI: 10.19595/j.cnki.1000-6753.tces.211562
电工理论与新技术 |
基于深度置信网络算法的面向铁磁材料旋转磁滞损耗的矢量磁滞模型
马阳阳1,2, 李永建1, 孙鹤1, 杨明1, 窦润田1
1.省部共建电工装备可靠性与智能化国家重点实验室(河北工业大学) 天津 300130;
2.国网河北省电力有限公司沧州供电分公司 沧州 061000
Vector Hysteresis Model for Rotational Hysteresis Loss of Ferromagnetic Materials Based on Deep Belief Network Algorithm
Ma Yangyang1,2, Li Yongjian1, Sun He1, Yang Ming1, Dou Runtian1
1. State Key Laboratory of Reliability and Intelligence of Electrical Equipment Hebei University of Technology Tianjin 300130 China;
2. State Gird Cangzhou Electric Power Supply Company Cangzhou 061000 China
全文: PDF (2000 KB)   HTML
输出: BibTeX | EndNote (RIS)      
摘要 铁磁材料磁滞建模是电气工程领域的基础性理论研究之一。该文基于深度置信网络(DBN)算法结合磁滞算子空间理论提出一种矢量磁滞模型。在模型结构中,引入郎之万函数作为映射函数对磁滞数据进行输入转换计算。利用多个磁滞算子构建算子空间生成高维算子数据,算子空间的数据输出作为DBN模型的输入,结合DBN算法表征算子数据与模型输出的非线性关系。利用样本的磁感应强度数据和生成的算子数据训练模型,获得模型参数。通过仿真表明构建的模型可以有效地描述铁磁材料在旋转磁化情况下的非线性特性和各项异性。同时,结合磁损分离理论改进磁损模型中相应的损耗系数,构建动态磁损计算模型,并将磁滞模型获得的数据应用于动态损耗计算。仿真表明,构建的磁滞模型可以有效地表征铁磁材料的实际磁化特性和损耗情况。
服务
把本文推荐给朋友
加入我的书架
加入引用管理器
E-mail Alert
RSS
作者相关文章
马阳阳
李永建
孙鹤
杨明
窦润田
关键词 磁滞模型深度置信网络算法磁滞算子磁滞损耗    
Abstract:The silicon steel sheet is the core material of electrical equipment, and its magnetization characteristics directly affect the operation mechanism of equipment. So, the hysteresis modeling of ferromagnetic materials is one of the basic theoretical studies in the field of electrical engineering. In this paper, a vector hysteresis model is proposed based on the deep belief network (DBN) algorithm and hysteresis operator space theory.
The structure of the model consists of three parts: input mapping function, operator space and DBN model. In this paper, the Langevin function is used as the input mapping function to calculate the input mapping of hysteresis data, so that the data can adapt to the characteristics of hysteresis operator in the subsequent structure and can reflect the saturation characteristics of hysteresis phenomenon. Hysteresis operators in multiple directions in H space construct a hysteresis operator space. And the magnetization trajectory of the material mapped by Langevin function is projected in all directions on H space. The high-dimensional hysteresis operator data is generated by calculating hysteresis operators in all directions. Then the output of the operator space is taken as the input of the DBN model. In the construction of vector hysteresis model, DBN model is mainly used to characterize the nonlinear relationship between the high-dimensional vector data output by the operator and the magnetic induction data of the material. The parameters of the vector hysteresis model are obtained by training the magnetic induction data of training samples and the operator data generated by the training samples. The model parameters are mainly obtained by training DBN parameters. And the training process of DBN mainly consists of two parts: (1) The CD algorithm is used to the pre-training of the RBM in each layer, then the RBMs are stacked to obtain the preliminary optimization parameters of the model. (2) The parameters obtained by pre-training are taken as initial values, and the Nadam optimizer is used for global parameter tuning to obtain the final optimization parameters of the model. The obtained model is fitted under the conditions of high, middle and low magnetic density (Bm=0.5 T, 1 T, 1.5 T) respectively, and it is proved that the trajectory error between the calculated data of the model and the original magnetization data is small. In addition, the x-axis and y-axis decomposition of the calculated vector hysteresis data also prove that the proposed vector model has a relatively small error in terms of phase for hysteresis data, thus ensuring the reliability of magnetic loss calculation. So, the simulation results of hysteresis data obtained by experiment show that the model can effectively describe the nonlinear characteristics and anisotropic of ferromagnetic materials under the rotation vector excitation.
Based on the magnetic loss separation theory, an improved loss calculation model is proposed in this paper. In the magnetic loss calculation model, the deviation between the magnetic loss calculated by the data calculated by the model and the actual magnetic loss is checked by the conversion function. Thus, the independence of the magnetic loss calculation and the characterization of the magnetic characteristics of the model are effectively guaranteed. And the data obtained by the hysteresis model is applied to the dynamic loss calculation. The simulation results show that the hysteresis model can fit the actual situation effectively.
Key wordsHysteresis model    deep belief network (DBN) algorithm    hysteresis operator    hysteresis loss   
收稿日期: 2021-05-14     
PACS: TM15  
基金资助:国家自然科学基金重点项目(52130710)、国家自然科学基金项目(51777055, 51977122)和河北省自然科学基金创新群体项目(E2020202142)资助
通讯作者: 李永建 男,1978年生,教授,博士生导师,研究方向为工程电磁场与磁技术、三维磁特性测量与建模。E-mail:liyongjian@hebut.edu.cn   
作者简介: 马阳阳 男,1991年生,博士研究生,研究方向为电磁场理论、铁磁材料磁滞建模与深度学习。E-mail:1367767122@qq.com
引用本文:   
马阳阳, 李永建, 孙鹤, 杨明, 窦润田. 基于深度置信网络算法的面向铁磁材料旋转磁滞损耗的矢量磁滞模型[J]. 电工技术学报, 2023, 38(15): 4063-4075. Ma Yangyang, Li Yongjian, Sun He, Yang Ming, Dou Runtian. Vector Hysteresis Model for Rotational Hysteresis Loss of Ferromagnetic Materials Based on Deep Belief Network Algorithm. Transactions of China Electrotechnical Society, 2023, 38(15): 4063-4075.
链接本文:  
https://dgjsxb.ces-transaction.com/CN/10.19595/j.cnki.1000-6753.tces.211562          https://dgjsxb.ces-transaction.com/CN/Y2023/V38/I15/4063