Abstract: To achieve collision-free and efficient operation of the multi-robot collaborative machining system for handling the complex structures while the tasks had temporal and ordering constraints,a Temporal and Ordering Constraint (TOC) modeling method and an Iterative Local Search (ILS) algorithm combined with insertion heuristics were proposed.The synchronization and precedence constraints among the tasks were modeled via a directed graph,and constraint consistency was verified by a modified Depth-First Search (DFS).An insertion heuristic was used to rapidly insert unconstrained tasks into the robot schedule,so as to form a complete and feasible initial solution.To optimize the operation efficiency,an iterative local search strategy was proposed to drive the directed graph model and optimize the uncertain part of the constraint set,so that the optimal schedule could be obtained.Simulation results showed that the proposed method could effectively improve the multi-robot processing efficiency by more than 14% compared with the current applied method.The reliability of the collision avoidance was verified by real-world robots,which showed batter operation safety of the planning results.

Key words: multi-robot cooperative machining, temporal and ordering constraints, collision-free scheduling, task allocation

摘要： 针对多机器人协同制造系统处理复杂产品时任务的时序约束复杂、避碰和效率优化目标难实现的问题,提出一种时序约束(TOC)建模方法与结合插入启发式的迭代局部搜索(ILS)算法。借助有向图建模任务的并发约束及优先约束,并通过改进深度优先搜索实现约束一致性检验;采用插入启发式实现无约束任务向机器人调度计划的快速插入,从而形成完整可行的初始解;为实现加工效率优化目标,提出采用迭代局部搜索策略驱动有向图模型对约束集不确定部分进行寻优,从而得到效率最优加工调度。仿真结果显示,所提出的方法相比目前工程应用方法有效提升多机器人加工调度效率17%以上,多机器人避碰可靠性则通过实机试验平台进行验证,显示规划结果具有良好的防碰撞安全性。

关键词: 多机器人协同制造, 时序约束, 无冲突调度, 任务分配