基于多机自律协同的柔性作业移动机器人调度

Flexible job shop scheduling based on multiple mobile robots autonomous collaboration

针对柔性作业多移动机器人调度面临的任务分配难以满足多属性需求、路径规划难以解决多机无冲突协同作业等问题,提出了基于多机自律协同的调度方法.一方面,根据自动换料、充电等常规任务与非均衡突发任务约束,提出了基于机床、任务和多机状态域的动态加权指派法,实现了多属性自趋优的灵活应变机制;另一方面,考虑位置、时间双维度冲突特征,设计了基于离散时间步的带权有向流网络,利用改进A*寻优算法完成了无碰撞路径搜索,继而设计了基于离散时间窗的在线冲突检测策略.测试结果表明:所提出的动态加权指派法具有更优越的任务分配综合评价性能,同时改进的路径规划算法能解决多移动机器人路径冲突检测问题.

A scheduling method based on the autonomous collaboration of multiple mobile robots was proposed to achieve task assignment with multi-attribute demand and path planning for achieving conflict-free cooperative operation.Based on the machine tool state domain,task state domain and multi-machine state domain,the dynamic weighted assignment method was studied according to the constraints of conventional tasks.The constraints include automatic loading and charging and the non-equilibrium burst task.In this way,the flexible strain mechanism with multi-attribute self-optimization can be realized.Considering the conflict characteristics of spatial position and moving time,a weighted extended flow network was designed on the basis of discrete time steps.Then,the improved A* optimization algorithm with an online conflict detection strategy was proposed to realize the non-collision path search.The test results show that the dynamic weighted assignment method has better performance for task allocation and comprehensive evaluation.Meanwhile,the enhanced path planning algorithm can effectively solve the path conflict detection issue.