二维电磁场-流体-温度场耦合仿真节点数据映射算法研究  

Research on Node Data Mapping Algorithm for the 2D Coupling Electromagnetic-Fluid-Thermal Fields

作　　者：刘刚[1] 张瀚方 池骋[1] 李琳[3] 李慧奇[1] 

机构地区：[1]华北电力大学河北省输变电设备安全防御重点实验室,保定071003 [2]中国电力科学研究院有限公司,北京100192 [3]新能源电力系统国家重点实验室(华北电力大学),北京102206

出　　处：《电工技术学报》2018年第1期148-157,共10页Transactions of China Electrotechnical Society

基　　金：国家自然科学基金(51407075);河北省自然科学基金(E2015502004);中央高校基本科研业务费专项资金(2015MS79)资助项目

摘　　要：在二维电磁场-流体-温度场耦合分析中,针对电磁场计算域与流体-温度场仿真域网格不匹配的情况,需要引入异构网格节点数据映射算法来实现数据的传递。为将电磁场损耗准确映射到流体-温度场的相关网格单元和节点上,提出快速映射法和径向基函数法,并结合两种方法的优点提出异构网格节点数据映射的混合算法。结合具有解析解的算例和换流变压器铁心漏磁场映射的实例,验证三种算法的可行性;讨论网格疏密对映射算法精度的影响;分析比较径向基函数法、快速映射法和混合算法的映射精度,并从计算效率方面对径向基函数法和快速映射法进行对比。结果表明:快速映射法、径向基函数法和混合算法各具优缺点,对不同算例,需要根据其模型和网格特点选择适当的映射算法,以满足精度要求,为二维电磁场-流体-温度场耦合仿真奠定了基础。In the analysis of the 2D coupling electromagnetic-fluid-thermal fields, which the meshes in electromagnetic domain don＇t match the meshes in the fluid-thermal domain, the mapping method of the node data need to be studied. In order to map electromagnetic losses to fluid-thermal field accurately, this paper proposed the fast projection algorithm and the radial basis function （RBF） scheme, and presented the mixed method combing the advantages of the two proposed methods. Based on the examples with analytical solutions and the leakage magnetic field mapping of a converter transformer, the feasibility of the three methods were verified, the effects of the grid density on the mapping algorithm accuracy was discussed, the mapping accuracy of the three methods were compared, and thecomputational efficiency of the fast projection algorithm and the RBF schemes were studied. The result shows that three methods have advantages and disadvantages. For different applications, it is necessary to choose a proper mapping algorithm according to the model and grid features. Above research lay a foundation for the 2D coupling electromagnetic-fluid-thermal fields.

关 键 词：数据映射 快速映射法 径向基函数法 混合算法 

分 类 号：TM401[电气工程—电器]