基于智能算法的感应电机多目标优化设计

Multi-objective Optimization Design of Induction Motor Based on Intelligence Algorithm

感应电机自身作为一个复杂系统,其设计变量多,多个设计目标之间相互约束,多目标优化设计过程复杂、限制因素多。针对这种情况,以一台24 V低压大电流感应电机为例,选取了定转子槽6个相关设计参数作为优化变量,以感应电机3个外特性参数(最大转矩、启动电流、效率)作为优化目标,运用计算机辅助电机设计软件ANSYS Maxwell得到了大量工程实验数据,利用BP神经网络对感应电机数学模型进行拟合,并用遗传算法寻找设计变量最优解,使用神经网络的预测功能寻找优化目标最优解。最后用ANSYS Maxwell将最优解进行工程验证,实现了感应电机的多目标优化设计。

As a complex system,induction motor has many design variables,and many design objectives restrict each other.Its multi-objective optimization design process is complex and has many restrictive factors.Taking a 24 V low voltage and high current induction motor as an example,six design parameters of stator and rotor slots are selected as optimization variables,and three external characteristic parameters(maximum torque,starting current and efficiency)of induction motor are taken as optimization objectives.Firstly,a large number of engineering experimental data are obtained by using computer aided motor design software ANSYS Maxwell.The mathematical model of induction motor is fitted by BP neural network,and the optimal solution of design variables is searched by genetic algorithm.The optimal solution of optimization target is searched by the prediction function of neural network.Finally,ANSYS Maxwell is used to validate the optimal solution and realize the multi-objective optimization design of induction motor.