基于二代非支配排序遗传算法的电子变压器多目标优化

Optimization design of high-power electronic transformer based on NASG-Ⅱ

基于二代非支配排序遗传算法(NSGA-Ⅱ)对大功率电子变压器的优化设计进行了研究。以电子变压器的面积乘积、磁芯损耗和绕组损耗为目标函数建立了三维优化模型。通过对NSGA-Ⅱ算法的研究,合理确定了选择、交叉、变异等算子的取值,对冷轧硅钢片、铁基非晶合金、超微晶合金三种磁芯材料的电子变压器的设计进行了优化。根据磁芯损耗和绕组损耗相等时变压器效率最大的原则选取了最优点。通过对优化结果的参数指标比较,分析了三种磁芯材料的物理特性。利用有限元软件计算了三种变压器的温度场,作了温度校核,并对三种变压器的热特性进行了分析,验证了超微晶合金材料在大功率电子变压器设计制造的优越性。

Optimization design of high-power electronic transformer is researched based on NASG-Ⅱ in this paper. Three-dimension model is established by adopted area product, loss of core, and loss of winding as the object func-tions. Optimization designs about the three kinds of electronic transformer, which the cores are silicon steel, amor-phous, and nanocrystalline, are carried on. The three factors of selection, crossover, and mutation are properly deter- mined. The optimum point is chosen by the principle that the efficiency of transformer is the highest when the 10ss of core equals to the loss of winding. The physical characters of the three transformers is analyzed according to the results of optimization design. The temperature checks are made by numerical calculation.