基于改进NSGA-Ⅱ的干式变压器优化设计

Optimal Design of Dry-type Transformer Based on Improved NSGA-Ⅱ

针对现存的干式变压器设计方法存在设计周期长、设计成本高和效率低等问题,提出了一种基于改进快速非支配多目标遗传算法(NSGA-Ⅱ)优化设计方法。首先,根据变压器设计中电磁理论和设计原则,给出了干式变压器的设计流程;其次,分析了干式变压器决策变量选择和温升参数等约束条件,给出了基于NSGA-Ⅱ算法优化数学模型;最后,改进了NSGA-Ⅱ算法,采用变量分组策略,完成干式变压器优化设计任务。对一台三相干式变压器进行优化设计,与人工设计方法相比,总成本降低了25.24%,总损耗降低了19.54%,而且收敛速度比传统NSGA-Ⅱ快,证实了该方法的有效性和可行性。

In order to solve the problems of long design cycle,high design cost and low efficiency in the existing design methods of dry-type transformer,an optimization design method based on an improved fast non-dominated multi-objective genetic algorithm(NSGA-Ⅱ) is proposed in this paper.Firstly,according to the electromagnetic theory and design principle in transformer design,the design process of dry-type transformer is given.Secondly,the constraints such as the selection of decision-making variables and temperature-rise parameters for dry-type transformer are analyzed,and an optimization mathematical model based on NSGA-Ⅱ algorithm is given.Finally,the NSGA-Ⅱ algorithm is improved,and variable grouping strategy are used to complete the optimal design task of dry-type transformer.The experiment optimizes the design of a three-phase dry-type transformer.Compared with the manual design method,the total cost is reduced by 25.24%,the total loss is reduced by 19.54%,and the convergence speed is faster than that of the traditional NSGA-Ⅱ.