【目的】针对核反应堆控制棒用驱动电机面临的大气隙结构导致的磁性能衰减、高温环境温升限制及多目标优化复杂等问题，本文提出了一种兼顾高效性与精度的优化设计方法，以提升电机的输出转矩、运行平稳性及热可靠性。【方法】首先，对电机进行电磁方案设计，确定24槽4极的极槽配合、径向“一”字型转子磁路及Recoma28钐钴永磁体材料的初始方案；然后，选取7项关键结构参数作为优化变量，基于最大最小拉丁超立方采样获取样本数据，通过Pearson相关系数分析参数灵敏度，构建高精度Kriging代理模型替代有限元仿真；其次，采用非支配排序遗传算法II(NSGA-II)开展以平均转矩最大化、转矩脉动最小化为目标的多目标优化；最后，通过Fluent软件进行温度场仿真，校核所设计的冷却结构的有效性。【结果】优化后电机平均转矩为17.074 N·m，较初始方案提升15.03%；转矩脉动降至0.29%，降幅达87.97%；额定工况下最高温度为117.8 ℃，温升58 K，满足F级绝缘与B级温升限值要求。【结论】Kriging代理模型与NSGA-II的结合有效解决了多参数优化的高计算复杂度问题，所设计的电机在满足结构与环境约束的前提下，电磁性能与热稳定性显著提升，可为核反应堆控制棒驱动系统提供可靠的动力支撑。

[Objective] To address the challenges faced by drive motors for nuclear reactor control rods, including magnetic performance attenuation caused by the large air-gap structure, temperature rise limitation in high-temperature environments, and the complexity of multi-objective optimization, this paper proposes an optimization design method that balance efficiency and accuracy. This approach aims to enhance the output torque, operational stability, and thermal reliability of the motor. [Methods] Firstly, an electromagnetic design scheme was developed for the motor, determining the initial scheme of 24 slots and 4 poles pole-slot combination, radial "I" type rotor magnetic circuit and Recoma28 samarium-cobalt permanent magnet material. Secondly, seven key structural parameters were selected as optimization variables. Sample data were obtained based on maximum-minimum Latin hypercube sampling. Parameter sensitivity was analyzed using Pearson correlation coefficient, and a high-precision Kriging surrogate model was constructed to replace finite element simulation. Thirdly, the non-dominated sorting genetic algorithm II (NSGA-II) was employed to conduct multi-objective optimization targeting maximization of average torque and minimization of torque ripple. Finally, temperature field simulations were performed using Fluent software to validate the effectiveness of the designed cooling structure. [Results] After optimization, the average torque of the motor reached 17.074 N·m, which was 15.03% higher than the initial scheme, the torque ripple was reduced to 0.29%, with a decrease of 87.97%. The maximum temperature under rated working conditions was 117.8 ℃, and the temperature rise was 58 K, meeting the requirements of class F insulation and class B temperature rise limit. [Conclusion] The combination of Kriging surrogate model and NSGA-II effectively solves the problem of high computational complexity of multi-parameter optimization. On the premise of meeting structural and environmental constraints, the designed motor has significantly improved electromagnetic performance and thermal stability, which can provide reliable power support for the nuclear reactor control rod drive system.

辽宁省自然科学基金联合基金(2023-MSLH-245)