首页 | 本学科首页   官方微博 | 高级检索  
     检索      

无轴轮缘推进器电磁驱动结构优化设计
引用本文:鲁浩然,张健,吕传茂,潘康,李政民卿.无轴轮缘推进器电磁驱动结构优化设计[J].科学技术与工程,2024,24(8):3236-3242.
作者姓名:鲁浩然  张健  吕传茂  潘康  李政民卿
作者单位:南京航空航天大学机电学院;南京航空航天大学机电学院;南京航空航天大学江苏省精密与微细制造技术重点实验室
基金项目:江苏省自然科学基金青年科学(BK20200424);国家自然科学基金资助项目(52105061); 2020国家重点研发计划(2020YFB2008100)
摘    要:无轴轮缘推进器对中国水下事业有重要作用,其电磁驱动部分采用大内径内转子永磁无刷直流电机结构。对无轴轮缘推进器电磁驱动部分的永磁无刷直流电机结构进行设计,并基于Maxwell进行电机性能优化分析,以提高其性能。选取效率和起动电流等4个参数为优化目标,选取与槽型相关的6个参数为优化变量。采用反向传播(back propagation, BP)神经网络建立优化变量与优化目标的关系式,采用线性加权法确定适应度函数,对电机进行多目标优化分析以提高其性能。最后将求得的最优解代回到Maxwell电机模型,验证求解结果的有效性。结果表明:优化后电机各项性能有所提升,优化结果较好,达到了预期效果。

关 键 词:结构优化  神经网络  线性加权法  多目标优化
收稿时间:2023/5/25 0:00:00
修稿时间:2024/3/14 0:00:00

Optimization Design of Electromagnetic Drive Structure for Shaftless Rim Thruster
Lu Haoran,Zhang Jian,Lu Chuanmao,Pan Kang,Li Zhengminqing.Optimization Design of Electromagnetic Drive Structure for Shaftless Rim Thruster[J].Science Technology and Engineering,2024,24(8):3236-3242.
Authors:Lu Haoran  Zhang Jian  Lu Chuanmao  Pan Kang  Li Zhengminqing
Institution:School of Mechanical and Electrical Engineering,Nanjing University of Aeronautics and Astronautics
Abstract:The shaftless rim thruster has an important role underwater construction in China, and its electromagnetic drive part uses a large diameter inner rotor permanent magnet brushless DC motor structure. In this paper, the permanent magnet brushless DC motor structure of the electromagnetic drive part of the shaftless rim thruster was designed and the motor performance optimization analysis was carried out based on Maxwell to improve its performance. Four parameters such as efficiency and starting current were selected as the optimization objectives, and six parameters related to the slot type were selected as the optimization variables. The BP neural network was used to establish the relationship between the optimization variables and the optimization target. The linear weighting method was used to determine the fitness function to perform a multi-objective optimization analysis of the motor to improve its performance. Finally, the optimal solution was substituted back to the Maxwell model to verify the validity of the solution results. The results showed that the performance of the motor was improved after optimization, and the optimization results were fine and achieved the expected results.
Keywords:Structural optimization    Neural network    Linear weighting method    Multi-objective optimization
点击此处可从《科学技术与工程》浏览原始摘要信息
点击此处可从《科学技术与工程》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号