• CN:11-2187/TH
  • ISSN:0577-6686

机械工程学报 ›› 2021, Vol. 57 ›› Issue (8): 65-80.doi: 10.3901/JME.2021.08.065

• 特邀专栏:庆祝厦门大学机电工程系建系80周年:微纳制造与智能制造 • 上一篇    下一篇

扫码分享

基于运行数据驱动反向设计的复杂装备个性化定制

侯亮, 林浩菁, 王少杰, 连晓振, 张炜   

  1. 厦门大学机电工程系 厦门 361102
  • 收稿日期:2020-11-27 修回日期:2021-03-20 出版日期:2021-04-20 发布日期:2021-06-15
  • 通讯作者: 侯亮(通信作者),男,1974年出生,博士,教授,博士研究生导师。主要研究方向为产品大批量定制技术、振动噪声控制以及工业大数据等。Email:hliang@xmu.edu.cn
  • 作者简介:林浩菁,女,1989年出生,博士研究生。主要研究方向为工业大数据分析、反向设计理论等。E-mail:394770523@qq.com;王少杰,男,1985年出生,博士,助理教授,硕士研究生导师。主要研究方向为工业大数据分析与人工智能系统研发等。E-mail:wsj@xmu.edu.cn;连晓振,男,1990年出生,博士研究生。主要研究方向为产品大批量定制技术、反向设计理论、工业大数据分析等。E-mail:xzhen_l@163.com;张炜,男,1982年出生,博士研究生。主要研究方向为大批量定制设计、生产线规划。E-mail:zhangw@stu.xmu.edu.cn
  • 基金资助:
    国家自然科学基金(51975495)、国家重点研发计划(2020YFB1709901)、创新方法工作专项(2020IM010100)和福建省中央引导地方科技发展专项(2020L3002)资助项目。

Mass Personalization for Complex Equipment Based on Operating Data-driven Inverse Design

HOU Liang, LIN Haojing, WANG Shaojie, LIAN Xiaozhen, ZHANG Wei   

  1. Department of Mechanical and Electrical Engineering, Xiamen University, Xiamen 361102
  • Received:2020-11-27 Revised:2021-03-20 Online:2021-04-20 Published:2021-06-15

摘要: 不同于消费类产品侧重外观配置等的显性个性化定制,装备产品往往需要在各种复杂环境和工况下高效、节能地完成预定任务。因此,装备产品的个性化定制不仅包含显性需求,还包含与使用场景或工况和运行性能与效率等相关的隐性个性化需求,后者对于提升产品竞争力具有更为重要的意义。在分析装备产品研发和运行特点及其面临的大数据机遇基础上,提出了一种基于运行数据驱动反向设计(Data-driven inverse design,DID)的复杂装备个性化定制方法(DID-MP)。首先给出了DID基本原理模型,并对比传统正向设计等方法分析了反向设计的基本特征。其次,基于传统大批量定制和运行数据驱动的反向设计,提出了一次定制和二次定制的概念,并给出了DID-MP的基本流程,阐述了反向设计目标选择、运行大数据采集、表征个性化使用环境的系统参数识别与建模,以及最优系统参数识别与前馈应用等关键技术。最后,给出了基于DID-MP的装载机变速箱模块优化定制以及动力传动控制系统定制升级的案例,并与传统设计结果对比表明DID-MP方法对于实现复杂装备个性化定制具有参考价值和意义。

关键词: 运行数据, 复杂装备, 反向设计, 个性化定制, 装载机

Abstract: Unlike consumer products with explicit personalization such as appearances, equipment products must finish scheduled tasks with energy efficiency in various complex usage scenarios and working conditions. Therefore, besides explicit personalized demands, equipment products contain implicit personalized demands hidden in diverse usage scenarios or working conditions, which challenges the personalization of equipment products. By analyzing the R&;D, operating characteristics, and big-data opportunities of equipment products, a mass personalization method based on data-driven inverse design (DID-MP) is proposed for complex equipment. Firstly, the data-driven inverse design is compared with the traditional forward design process, in order to introduce and highlight data-driven inverse design (DID) characteristics. Secondly, based on DID and traditional mass customization, two concepts, i.e., first-time customization and second-time customization, are proposed. The DID-MP framework for complex equipment is given together with the key techniques for objective selection, operating data acquisition, identification and modeling of system parameters based on usage scenarios, and optimal system parameter identification and feedforward application. Finally, a case study of optimizing and updating the loader transmission and control module is used to verify the proposed mass personalization method using DID. The comparison with the traditional methods shows that the proposed method is powerful to realize mass personalization of complex equipment.

Key words: operating data, complex equipment, inverse design, mass personalization, loaders

中图分类号: