摘要
中空型行波超声电机柔性转子由于结构复杂,因此存在设计空间较小、转子与定子接触界面接触不均匀、转子变形后内应力较大等问题。该文首先利用有限元方法建立转子尺寸参数化有限元模型;然后用拉丁超立方(LatinHypercube)随机抽样法,在设计空间内建立响应面模型;最后,利用自适应遗传优化算法(self-adaptive genetic algorithm,SAGA)基于响应面模型得到设计空间内的全局最优解。为了验证优化结果的合理性,设计了定子外径为ф70 mm,陶瓷片内外径尺寸分别为ф55 mm和ф70 mm的超声电机配合该文设计的转子进行磨合实验。实验证明,经过优化设计得到的转子与定子之间接触更均匀,转子变形后内应力减小,转子变形空间更大。
The small design space, uneven contact with the contact interface of the rotor and the stator, and larger rotor deformation stress are problems of the rotor structure of a hollow traveling wave ultrasonic motor. Firstly, parametric finite element model of the rotor was established. Then response surface model was established in the design space by the Latin-Hypercube sampling method. Finally, the global optimal solution within the space which is based on response surface model was got by self-adaptive genetic optimization design. In order to verify the reasonableness of the optimization results, this paper designed an ultrasonic motor. Its diameter of the stator was 70 mm. Its inside diameter size of the ceramic pieces was 55 mm. Its outside diameter size of the ceramic pieces was 70 mm. Experiments show that the contact between the rotor and the stator was more uniform, the rotor deformation stress was reduced, and the rotor deformation space was greater, after optimization.
出处
《中国电机工程学报》
EI
CSCD
北大核心
2014年第30期5378-5385,共8页
Proceedings of the CSEE
基金
国家自然科学基金项目(51175264
51275235)
中央高校基本科研业务费专项资金资助项目(NJ20130004
NJ20140026)~~
关键词
超声电机
响应面模型
自适应遗传算法
拉丁超立方抽样
结构优化
ultrasonic motor
response surface methodology（RSM）
self-adaptive genetic algorithm（SAGA）
Latin-Hypercube sampling
structural optimization
