基于多约束的焊接机器人任务分配与路径规划研究

Welding robbots have been widely used in the assembly for body-in-white(BIW). Multiple robots welding operation face the problem of the interference in overlapping working area, uneven task assignments，unreasonable welding paths, and so on. To solve the above problems, the research on task assignments and path planning for welding robots is carried out. Considering the production cycle time, the welding distribution and number in BIW, a model for predicting the number of welding robots is established based on the fuzzy theory. Considering the coverage of robots, task uniformity and mutual non-interference principle, a mathematical model for task allocation based on dynamic programming is established. The welding points in overlapping area are allocated dynamically to robots one by one, which ensure the feasibility of the task. In order to avoid the collision and the delay, a strategy of parallel search for welding path is proposed. Considering the welding sequence constraints and the collision free, the mathematical model of the robot path optimization is established. The directed graph is used to represent the possible welding sequence. The nearest neighbor matrix for welding point is built to generate the initial population for genetic algorithm, which can accelerate the speed for of the genetic algorithm and avoid searching local optimal solution. By optimizing the welding robot welding task allocation and path, the efficiency and quality of welding BIW welding robot can be improved greatly. In addition, the above proposed theories and methods can be applied in other types of robots.

焊接机器人已被广泛应用在白车身装配生产线上，但多个机器人焊接作业会出现交叉作业区干涉、任务分配不均、焊接路径不合理等问题。针对以上出现的问题，研究焊接机器人任务分配和焊接路径规划的理论和方法。考虑生产周期及焊点多少和焊点分布情况，建立基于模糊理论机器人数量的预估模型。考虑机器人的所达范围、任务均匀性原则，建立基于动态规划的任务分配的数学模型，分阶段对共同作业区焊点动态分配，保证分配方案的可行性。为了避免碰撞和等待，研究基于多机器人路径并行搜索的策略，实现多机器人的避障与最优协调。考虑焊接顺序和避让策略，建立基于有向图的路径优化的数学模型；通过焊点间最邻近点矩阵为遗传算法提供初始种群，以加快求解速度。通过对焊接机器人任务分配和焊接路径优化理论和方法的研究，提高白车身焊接机器人的效率和焊接质量。同时，这些理论和方法也可为其它类型的机器人协调和路径规划提供理论参考价值。

本课题针对白车身生产线上多焊接机器人作业环境复杂，任务协调困难，焊接约束多等条件，围绕多机器人的协调、避障、路径优化和生产线布局展开研究。在理论方面，提出了基于动态路径修改的多机器人协调方法，通过评估动态路径修改成本确定最优的机器人路径修改顺序。提出了改进A*算法搜索路径，根据障碍物的大小和位置分别，动态修改A*算法的搜索步长，大大提高了避障路径搜索的速度。提出三次B样条优化已搜索的机器人路径，使机器人的速度、加速度和加加速度都得到了平滑，运动更加平稳，并减少对机器人关节的冲击。此外，对生产线布局进行了优化，提出了平行装配线的多目标优化方案，为多机器人在生产线上合理布局、高效协作完成生产任务提供新的解决方案。在实验方面，搭建了以UR5机器人为基础的机器人避障实验平台，通过机器人操作系统(ROS)实现对提出避障算法和路径优化算法进行仿真和实验。课题的研究为实现工业机器人高效的运动规划提供了进一步的理论的基础。同时，课题的研究为进一步展开对机器人高效避障，特别是考虑机器人各个关节自身所占空间作为障碍物的避障；机器人路径平滑性优化奠定了理论和实验基础。在国内外期刊上发表文字9篇，其中1篇SCI收录，1篇EI期刊论文，3篇会议论文均被EI收录，其余均为中文核心期刊。