基于粒子群算法的核屏蔽感应电机多目标优化设计

Multi-Objective Optimization Design of Nuclear Canned Induction Motor Based on Particle Swarm Optimization

屏蔽感应电机是核反应堆冷却回路的动力源,提高该电机的效率和降低电磁噪声具有重要的意义。将有限元法和磁路法相结合,设计了1台20kW核屏蔽感应电机。利用有限元软件对其进行额定工况运行状态下噪声和损耗仿真分析,确定影响电机效率和电磁噪声的主要参数;选取影响电机效率和噪声的主要参数为优化变量,电机电磁噪声和效率为优化目标,输出功率为约束条件;采用正交试验表获得样本空间;通过响应面法建立优化目标及约束条件的优化模型,最后利用粒子群算法对优化模型进行求解。有限元仿真结果表明:优化后电机效率提高了2.6%,电磁噪声降低了5.2dB。

Canned induction motor was the power source of nuclear reactor cooling circuit.It was significant to improve the efficiency and reduce the electromagnetic noise of the motor.A 20 kW nuclear canned induction motor was designed by combining the finite element method with magnetic circuit method.The main parameters that affected the motor efficiency and electromagnetic noise were determined by the simulation analysis of noise and loss under the rated operating condition with the finite element software.The main parameters that affected the efficiency and noise of the motor were selected as optimization variables,the electromagnetic noise and efficiency of the motor were the optimization objectives,and output power was the constraint condition.Sample space was obtained by the orthogonal test table,the optimization model was obtained by response surface methodology,and the particle swarm optimization was used to solve the optimization model.The finite element simulation results showed that the efficiency of the optimized motor was improved by 2.6 % and the electromagnetic noise was reduced by 5.2 dB.