航空学报 > 2023, Vol. 44 Issue (7): 427162-427162   doi: 10.7527/S1000-6893.2022.27162

基于PSO与WSVD的飞机部件位姿拟合方法

何晓煦1(), 雷沛1,2, 潘登1, 杨阳1,2, 李现坤1,2, 邓珍波1   

  1. 1.航空工业成都飞机工业(集团)有限责任公司,成都  610092
    2.四川省航空智能制造装备工程技术研究中心,成都  610092
  • 收稿日期:2022-03-15 修回日期:2022-04-19 接受日期:2022-06-17 出版日期:2023-04-15 发布日期:2022-08-04
  • 通讯作者: 何晓煦 E-mail:auroraimelanie@126.com

Posture and position computing method for aircraft component based on PSO and WSVD

Xiaoxu HE1(), Pei LEI1,2, Deng PAN1, Yang YANG1,2, Xiankun LI1,2, Zhenbo DENG1   

  1. 1.Chengdu Aircraft Industrial (Group) Co. Ltd,Chengdu  610092,China
    2.Sichuan Aviation Intelligent Manufacturing Equipment Engineering Technology Research Center,Chengdu  610092,China
  • Received:2022-03-15 Revised:2022-04-19 Accepted:2022-06-17 Online:2023-04-15 Published:2022-08-04
  • Contact: Xiaoxu HE E-mail:auroraimelanie@126.com

摘要:

位姿拟合是飞机大部件调姿对合的核心环节之一,所采用的算法直接影响调姿对合的精度和效率。由于飞机部件不可避免地存在装配误差,并且不同区域调姿基准点的精度要求通常也存在差异,传统的奇异值分解法不能保证全部基准点都满足精度要求。加权奇异值分解法能够通过提高公差要求低的基准点的转换残差,来保证公差要求严格的基准点的转换精度,但权重如何准确分配仍缺乏有效的手段。针对以上问题,提出一种基于粒子群优化结合加权奇异值分解的位姿计算方法,利用粒子群优化算法搜索调姿基准点的权重值,保证每个基准点的转换残差均满足精度要求。通过在飞机数字化装配中的应用分析,验证了算法的有效性。

关键词: 飞机装配, 部件调姿, 位姿拟合, 加权奇异值分解, 粒子群优化

Abstract:

Posture and position computation is a key part for aligning and joining large aircraft component, which makes significant impacts on accuracy and efficiency. Due to the assembly errors of the aircraft components and the different accuracy requirements of the datum points, the traditional Singular Value Decomposition (SVD) method is usually invalid. Weighted Singular Value Decomposition (WSVD) method can improve the accuracy of the points with higher tolerance requirements by sacrificing the points with lower requirements. However, how to assign weights to the points is still a problem. A method based on Particle Swarm Optimization (PSO) and WSVD is proposed in this paper to solve this problem. The weights of the points is searched by PSO method to assure the transformation residue of each datum point meets its tolerance requirements. Through the application of aircraft digital assembly, the validity of the method was verified.

Key words: aircraft assembly, component alignment, posture and position computation, weighted singular value decomposition, particle swarm optimization

中图分类号: