[1]钟大鹏,万 立.基于几何特征推理与 PCA 导向修正的装配路径规划[J].机械与电子,2026,44(05):87-94.
　ZHONG Dapeng,WAN Li.Assembly Path Planning Based on Geometric Feature Inference and PCA-guided Correction[J].Machinery & Electronics,2026,44(05):87-94.
点击复制

基于几何特征推理与 PCA 导向修正的装配路径规划()
分享到：

《机械与电子》[ISSN:1001-2257/CN:52-1052/TH]

卷:
44
期数:
2026年05期
页码:
87-94
栏目:
智能制造
出版日期:
2026-05-27

文章信息/Info

Title:
Assembly Path Planning Based on Geometric Feature Inference and PCA-guided Correction
文章编号:
1001-2257 ( 2026 ) 05-0087-08
作者:
钟大鹏; 万 立
华中科技大学国家智能设计与数控技术创新中心,湖北 武汉 430074
Author(s):
ZHONG Dapeng ;  WAN Li
( National Center of Technology Innovation for Intelligent Design and Numerical Control , Huazhong University of Science and Technology , Wuhan 430074 , China )
关键词:
虚拟装配; 路径规划;  GIPC RRT ; 几何特征推理; 高斯管道采样; PCA 导向修正
Keywords:
virtual assembly ;  path planning ;  GIPC-RRT ;  geometric feature inference ;  Gaussian tube sampling ;  PCA-guided correction
分类号:
TP18 ;TG95
文献标志码:
A
摘要:
针对虚拟装配场景中严格几何配合约束导致采样失效、构型空间狭窄导致扩展效率低的问题,提出一种基于几何特征推理与 PCA 导向修正的路径规划算法( GIPC RRT )。首先,构建面向装配意图的几何推理机制,利用鲁棒主成分分析自动解析接触约束的法向分布,反演全局最优装配主轴;据此设计确定性高斯管道采样策略,通过构建沿主轴分布的各向异性概率密度场实现对装配可行域的高概率覆盖,有效解决轴孔配合等装配环境下的零测度采样难题。其次,提出基于增量 PCA 的导向修正扩展策略,利用局部路径演化趋势预测生长方向,并建立“预测—碰撞—修正”的闭环反馈机制,将碰撞信息转化为几何边界导引,实现对狭窄通道的柔顺探索。基于 NX 平台的仿真实验表明,该方法无需人工示教即可自动解析装配意图,在轴孔装配等装配场景下的规划成功率与计算效率显著优于传统算法,且具有较强的工程适应性。
Abstract:
Aiming at the problems of sampling failure caused by strict geometric mating constraints and low extension efficiency within narrow configuration spaces in virtual assembly scenarios , a path planning algorithm based on Geometric Feature Inference and PCA Guided Correction RRT ( GIPC-RRT ) is proposed.First , a geometric inference mechanism oriented towards assembly intent is constructed.By utilizing Robust Principal Component Analysis ( PCA ) to automatically parse the normal distribution of contact constraints , the global optimal assembly axis is inversely derived.Based on this , a deterministic Gaussian tube sampling strategy is designed.By constructing an anisotropic probability density field distributed along the principal axis , high probability coverage of the feasible assembly region is achieved , effectively solving the zero-measure sampling dilemma in assembly environments such as peg in hole mating.Second , a guided correction extension strategy based on incremental PCA is proposed.This strategy utilizes the local path evolution trend to predict the growth direction , and establishes a “ prediction collision cor- rection ” closed loop feedback mechanism.By transforming collision information into geometric boundary guidance , compliant exploration of narrow passages is realized.Simulation experiments conducted on the NX platform demonstrate that the proposed method can automatically parse assembly intent without manual teaching.The planning success rate and computational efficiency in assembly scenarios , such as peg-in-hole assembly are significantly superior to traditional algorithms , exhibiting robust engineering adaptability.

参考文献/References:

[ 1 ] 刘检华,孙清超,程晖 . 产品装配技术的研究现状、技术内涵及发展趋势[ J ] . 机械工程学报, 2018 ( 11 ): 16-42. [ 2 ] Tang Zhuozhen , Ma Zhuogzhong.An overview of path planning algorithms [ J ] .IOP Conference Series : Earth and Environmental Science , 2021 , 804 ( 2 ): 022024. [ 3 ] Xiong Jing , Hu Youmin , Wu Bo , et al.Minimum-cost rapid-growing random trees for segmented assembly path planning [ J ] .The International Journal of Advanced Manufacturing Technology , 2015 , 77 : 1043-1055. [ 4 ] Dalibard S , Laumond J.Linear dimensionality reduction in random motion planning [ J ] .International Journal of Rbotics Research , 2011 , 30 ( 12 ): 1461-1476. [ 5 ] 张璐璐,刘恩福,刘晓阳 . 一种装配路径规划改进 GoalBia-RRT 算 法 [ J ] . 机 械 设 计 与 制 造, 2018 ( 11 ):208-211. [ 6 ] 代俊杰 . 虚拟人双臂协作装配路径规划算法研究与实现[ D ] . 武汉:华中科技大学,2023. [ 7 ] 熊晶,段晓坤 . 狭窄空间内复杂产品的装配路径规划研究[ J ] . 现代制造技术与装备, 2021 , 57 ( 7 ): 11-14. [ 8 ] Yi Guodong , Zhou Chuanyuan , Cao Yanpeng , et al.Hybrid assembly path planning for complex products by reusing a priori data [ J ] .Mathematics , 2021 , 9 ( 4 ): 395. [ 9 ] Masehian E , Ghandi S.ASPPR : a new assembly sequence and path planner / replanner for monotone and nonmonot-one assembly planning [ J ] .Computer-Aided Design , 2020 , 123 : 102828. [ 10 ] Kuffner J J , Lavalle S M.RRT-connect : an efficient approach to single-query path planning [ C ]// IEEE International Conference on Robotics and Automation.New York : IEEE , 2000 : 995-1001. [ 11 ] 谭薪兴,李光,易静,等 . 改进 RRT 算法的机械臂路径规划[ J ] . 计算机集成制造系统, 2025 , 31 ( 3 ): 1014-1023. [ 12 ] 汪磊 . 虚拟装配中装配路径自动规划的研究与实现 [ D ] . 武汉:武汉理工大学,2018. [ 13 ] Issaoui L , Aifaoi N , Benamara A.Modelling and implementation of geometric and technological information for disassembly simulation in CAD environment [ J ] .The international Journal of Advanced Manufacturing Technology , 2017 , 89 ( 5 / 6 / 7 / 8 ):1731-1741. [ 14 ] 李思良,袁庆霓,胡涞,等 . 复杂装配体多层次装配路径规划研究[ J ] . 计算机仿真, 2020 , 37 ( 3 ): 178-182. [ 15 ] 胡伟,王晓楠 . 基于 APF-RRT*算法的装配机器人避障 路 径 规 划 [ J ] . 科 技 和 产 业, 2025 , 25 ( 17 ):73-78. [ 16 ] Xiao Guangzhou , Zhang Lixian , Wu Tong , et al.FBi RRT : a path planning algorithm for manipulators with heuristic node expansion [ J ] .Robotica , 2024 , 42( 3 ): 644-659.

相似文献/References:

[1]王 嵘,万永菁.一种基于ＳＬＡＭ 的多功能探索机器人设计[J].机械与电子,2019,(09):51.
　.Design of Multifunctional Exploration Robot Based on SLAM[J].Machinery & Electronics,2019,(05):51.
[2]谢智慧,卢道华,王 佳,等.基于改进蚁群算法的机器人路径规划问题研究[J].机械与电子,2019,(06):70.
　,,et al.Research on Robot Path Planning Problem Based on Improved Ant Colony Algorithm[J].Machinery & Electronics,2019,(05):70.
[3]关醒权,董文杰,莫鹏飞.双层立体车库结构设计及运动仿真分析[J].机械与电子,2019,(10):59.
　,Structural Design and Motion Simulation Analysis of Double-deck Stereo Garage[J].Machinery & Electronics,2019,(05):59.
[4]赵健,张阳.基于典型栅格地图的代价地图改进方法[J].机械与电子,2018,(12):73.
　ZHAO Jian,ZHANG Yang.Cost Map Improvement Method Based on Typical Grid Map[J].Machinery & Electronics,2018,(05):73.
[5]操松元１,陈 江２,严 波１,等.无人机巡检输电线路的路径规划算法研究[J].机械与电子,2019,(05):40.
　,,et al.Research on Path Planning Algorithms for Unmanned Aerial Vehicle Patrol Inspection Transmission Lines[J].Machinery & Electronics,2019,(05):40.
[6]文生平,张磊,刘其信.基于改进蚁群算法激光导航轮式机器人路径规划[J].机械与电子,2016,(05):73.
　WEN Shengping,ZHANG Lei,LIU Qixin.Path Planning for Laser Navigation Wheeled Robots Based on Improved Ant Colony Algorithm[J].Machinery & Electronics,2016,(05):73.
[7]周嵘,张志翔,翟晓晖,等.机器人室内路径规划算法的实用性研究[J].机械与电子,2016,(08):71.
　ZHOU Rong,ZHANG Zhixiang,ZHAI Xiaohui,et al.Practical Research on Robot Path Planning Algorithm[J].Machinery & Electronics,2016,(05):71.
[8]白金柯,吴晓娜.一种新型蚁群随机树的机器人路径规划算法[J].机械与电子,2015,(07):73.
　BAI Jinke,WU Xiaona.Robot-path Planning Based on Ant Colony Optimization and Rapidly-exploring Random Tree[J].Machinery & Electronics,2015,(05):73.
[9]李积云,许亚军,蒲卫华,等.基于势场法的路径规划算法特性分析和路径优化[J].机械与电子,2020,(05):18.
　,,et al. Characteristics Analysis of Algorithm and Methods of Optimizing Path Based on Artificial Potential Field Method[J].Machinery & Electronics,2020,(05):18.
[10]桑和成,宋栓军,唐铭伟,等.基于改进蚁群算法的机器人路径规划研究[J].机械与电子,2021,(02):17.
　Sang Hecheng,Song Shuanjun,Tang Mingwei,et al.Research on Robot Path Planning Based on Improved Ant Colony Algorithm[J].Machinery & Electronics,2021,(05):17.

备注/Memo

备注/Memo:
收稿日期: 2026-01-24 基金项目:湖北省中央引导地方科技发展专项( 2024BGA002 ) 作者简介:钟大鹏 ( 2001- ),男,江西新余人,硕士研究生,研究方向为装配路径规划;万 立 ( 1963- ),男,湖北武汉人,博士,教授,研究方向为机械设计及理论、数字化设计与制造和产品全生命周期管理。
更新日期/Last Update: 2026-08-25