改进量子遗传算法在电机结构优化中的应用

Application of Improved Quantum Genetic Algorithm to Optimization Design of Motor Structure

针对电机电磁设计过程中结构尺寸多、约束条件复杂的特点,以某旋转磁系统在外的永磁无刷双转电机为原型,建立了以电机质量为优化目标,包含10个优化变量和11个约束的数学模型,使用引进了动态旋转角控制策略、量子交叉操作、量子变异操作的改进量子遗传算法对模型进行搜索,获得了质量降低、约束指标满足要求的结构尺寸方案,并比较了在不同参数下搜索效果,可作为进一步改进算法的基础。

Considering the features of a large number of physical dimensions and complex constraints in electromagnetic design of a motor, a brushless DC motor（BLDCM） model with 10 physical dimensions and 11 constraints is built to optimize the quality of the counter-rotating BLDCM with an outer rotating magnetic system. The quantum genetic algo- rithm is improved by introducing dynamic adjustment mechanism of rotation angle of quantum gate, and quantum crossover and mutation operations to obtain a group of physical dimensions. Thus, an optimum structure design scheme is achieved with lighter weight of the motor, and the constraint indexes meet the requirements. In addition, the searching effects of different parameters are compared to provide a basis for further improving the algorithm.