汽车工程 ›› 2021, Vol. 43 ›› Issue (10): 1457-1465.doi: 10.19562/j.chinasae.qcgc.2021.10.006

• • 上一篇    下一篇

基于EGO加点策略的动力电池包多目标优化

王普毅1,3,白影春1,2(),林程1,武振江4,王保华3   

  1. 1.北京理工大学,电动车辆国家工程实验室,北京  100081
    2.湖南大学,汽车车身先进设计制造国家重点实验室,长沙  410082
    3.西北机电工程研究所,咸阳  712099
    4.中汽研(天津)汽车工程研究院有限公司,天津  300300
  • 收稿日期:2021-04-19 修回日期:2021-06-09 出版日期:2021-10-25 发布日期:2021-10-25
  • 通讯作者: 白影春 E-mail:baiyc@bit.edu.cn
  • 基金资助:
    国家自然科学基金(51805032);湖南大学汽车车身先进设计制造国家重点实验室开放基金(31915001)

Multi⁃objective Optimization of Traction Battery Pack Based on EGO Strategy with Additive Sample Points

Puyi Wang1,3,Yingchun Bai1,2(),Cheng Lin1,Zhenjiang Wu4,Baohua Wang3   

  1. 1.Beijing Institute of Technology,National Engineering Laboratory for Electric Vehicles,Beijing 100081
    2.Hunan University,State Key Laboratory of Advanced Design and Manufacturing for Vehicle Body,Changsha 410082
    3.Northwest Institute of Mechanical and Electrical Engineering,Xianyang 712099
    4.CATARC(Tianjin)Automotive Engineering Research Institute Co. ,Ltd. ,Tianjin 300300
  • Received:2021-04-19 Revised:2021-06-09 Online:2021-10-25 Published:2021-10-25
  • Contact: Yingchun Bai E-mail:baiyc@bit.edu.cn

摘要:

为了动力电池包的轻量化和提高其模态频率,提出一种基于EGO加点策略的多目标优化方法。首先通过实验设计和帕累托法则分析了设计变量对优化目标的影响,选出对电池包的质量和1阶模态频率影响较大的变量作为优化对象,以降低求解难度。其次采用MOPSO算法,辅以Kriging代理模型求解优化问题,再利用EGO加点策略和物理模型分别获得的新设计点和样本,进而更新代理模型直至优化收敛。最后,利用测试函数验证了所提方法的可行性,并将其用于随机振动下电池包多目标优化问题的求解。结果表明,该方法高效可行,在获得较高1阶模态频率的同时,电池包的质量减小了4.89 kg。

关键词: 电池包, 多目标优化, 代理模型, EGO加点策略

Abstract:

For the lightweighting of battery pack and increasing its modal frequency, a multi-objective optimization scheme based on efficient global optimization (EGO) strategy with additive sample points is proposed. Firstly, by using the design of experiment and Pareto rule, the effects of design variables on the optimization objectives are analyzed, and the variables having more significant influences on the mass and 1st-order modal frequency of battery pack are chosen to be optimized so as to reduce the problem-solving difficulty. Then, multi-objective particle swarm optimization (MOPSO) algorithm is adopted assisted with Kriging surrogate model to solve the optimization problem, and the EGO strategy with additive sample points is employed to get the new design points and samples respectively, with the surrogate model updated until the optimization procedure converges. Finally, the test functions are utilized to verify the effectiveness of the scheme proposed, which is then applied to the multi-objective optimization of battery pack. The results show the scheme is efficient and feasible, with which the mass of battery pack reduces by 4.89 kg while maintaining a higher 1st order modal frequency.

Key words: battery pack, multi?objective optimization, surrogate model, EGO with additive sample points