Computer Integrated Manufacturing System ›› 2024, Vol. 30 ›› Issue (7): 2389-2405.DOI: 10.13196/j.cims.2023.0040

Previous Articles     Next Articles

Path planning method of bidirectional extended manipulator under posture constraints

LIU Guanghui1,2+,GAO Jiahao1,2,MENG Yuebo1,2,XU Shengjun1,2,HAN Jiuqiang2   

  1. 1.School of Information and Control Engineering,Xi'an University of Architecture and Technology
    2.Xi'an Key Laboratory of Intelligent Technology for Building Manufacturing
  • Online:2024-07-31 Published:2024-08-07
  • Supported by:
    Project supported by the Key R&D  Program of Shaanxi Province,China(No.2021SF-429).

位姿约束下的双向扩展机械臂路径规划方法

刘光辉1,2+,高嘉豪1,2,孟月波1,2,徐胜军1,2,韩九强2   

  1. 1.西安建筑科技大学信息与控制工程学院
    2.西安市建筑制造智动化技术重点实验室
  • 作者简介:
    +刘光辉(1976-),男,陕西西安人,副教授,博士,研究方向:计算机视觉理解、建筑环境智能感知与调控、建筑智动化技术等,通讯作者,E-mail:guanghuil@163.com;

    高嘉豪(1995-),男,陕西宝鸡人,硕士研究生,研究方向:机器人智能规划算法、建筑机器人、计算机感知与理解等,E-mail:2008811558@xauat.edu.cn;

    孟月波(1979-),女,陕西西安人,教授,博士,研究方向:计算机视觉理解、建筑环境智能感知与调控、建筑智动化技术等,E-mail:mengyuebo@163.com;

    徐胜军(1976-),男,陕西西安人,副教授,博士,研究方向:视觉识别智能感知理论、人工智能与智动化系统、模型仿真智能控制理论等,E-mail:duplin@sina.com;

    韩九强(1951-),男,陕西西安人,教授,研究方向:智能测控理论与技术、智动化与机器人、智慧工厂等,E-mail:jqhan@mail.xjtu.edu.cn。
  • 基金资助:
    陕西省重点研发计划资助项目(2021SF-429)。

Abstract: Aiming at the problems of low planning efficiency,poor passing ability of link and the rough path in the path planning method of manipulator,based on the Rapidly-exploring Random Tree*(RRT*),a Bidirectional extended path planning of manipulator method under Posture Constraint and path Optimization named PCO-BT-RRT* was proposed.The Bidirectional extended RRT*(BT-RRT*)algorithm guided by target bias was designed,and the initialization process of the RRT* algorithm was improved.By taking the starting node and the target node as the initial nodes of the two random trees respectively,to improve the path generation speed while ensuring low path costs,the random trees were guided to grow in opposite directions with a certain probability through the target bias strategy,accelerating the exploration of unknown areas.A Posture Constraint and path Optimization strategy(PCO)was proposed,which used the kinematics model and collision detection rules of the manipulator to jointly constrain the expansion process of the new node,to find the obstacle avoidance path in the reachable space of the manipulator.The generated path was pruned by eliminating redundant nodes and shortening the length of feasible path,then by iteratively adjusting the position of path nodes to ensure the smooth path and improve the quality of path generation.The simulation results showed the significant effectiveness of the proposed method in path planning problems.The practicability of this method was verified by the real machine experiment on the self-developed BIM information fusion building masonry system.

Key words: path planning of manipulator, rapidly-exploring random tree* algorithm, bidirectional extended, posture constraints, path smoothing

摘要: 针对机械臂路径规划方法存在的规划效率低、连杆通过性差、路径粗糙等问题,以渐进最优快速随机搜索树RRT*为基础,提出一种位姿约束下的双向扩展机械臂路径规划方法(PCO-BT-RRT*)。首先,设计目标偏置引导的双向扩展RRT*算法(BT-RRT*),改进了RRT*算法的初始化过程,将起始点和目标点分别作为两棵随机树的初始节点,并通过目标偏置策略引导其以一定概率相向生长,加快探索未知区域,在保证路径代价较低的同时提升路径生成速度。其次,提出一种位姿约束路径优化策略(PCO),采用机械臂运动学模型和碰撞检测规则共同约束新节点扩展过程,寻找机械臂可达空间内的避障路径;对生成路径剪枝剔冗,缩短可行路径长度,同时对消冗节点以迭代调整的方式进行平滑优化,提高路径生成质量。通过仿真实验分析,验证了所提方法在路径规划问题上的显著性成效;在自主研发的BIM信息融合下建筑砌筑系统进行真机避障测试,验证了该方法的实用性。

关键词: 机械臂路径规划, RRT*算法, 双向扩展, 位姿约束, 路径平滑

CLC Number: