仓储物流机器人集群避障及协同路径规划方法

Obstacle avoidance and cooperative path planning method of warehouse logistics robot cluster

基于智能机器人的智能仓储系统为解决因电子商务兴起带来的仓储物流压力提供了有效的方案。而机器人集群的避障及路径规划问题是智能仓储系统能否正常运行以及提升其运行效率的关键所在。该文创新地提出一种在交通规则和预约表约束下的基于改进Q Learning算法的仓储物流机器人集群避障及协同路径规划方法。通过改进的Q Learning算法规划出每个机器人完成任务目标的最短路径并形成预约表,利用交通规则和预约表解决仓储物流机器人集群在运行时发生的碰撞和死锁问题,并根据所设定的协同机制,减少机器人无任务的待机状态,平衡各机器人之间的工作量,最终实现在保证系统安全运行的基础上缩短系统运行时间的目的。通过Matlab对该文所设计的算法进行仿真,以系统无碰撞完成所有任务的运行总时间即系统中最后一个完成任务的机器人无碰撞运行的总步数为评价指标,验证了该方法的有效性。

The intelligent warehousing system based on intelligent robot provides an effective solution for reduction of the warehousing logistics pressure caused by the rise of e commerce.The obstacle avoidance and path planning method of robot clus ter is the key to the normal operation of intelligent storage system and the improvement of its operation efficiency.An innovative approach based on improved Q Learning algorithm is proposed for obstacle avoidance and collaborative path planning of ware house logistics robot cluster under the constraints of traffic regulations and reservation form.The shortest path for each robot to complete its task goal is planned by means of the improved Q Learning algorithm,and a reservation table is formed.The traffic regulations and reservation table are used to solve the collision and deadlock problems of the warehouse logistics robot cluster while it is running.According to the established cooperative mechanism,the task free standby state of the robot is reduced,the workload of each robot is balanced,and the running time of the system on the basis of ensuring the safe running of the system is finally shortened.The algorithm designed in this paper is simulated by Matlab,the evaluation index is the total running time of the system to complete all tasks without collision,that is,the total number of steps of the last robot to complete tasks without collision,by which the effectiveness of the method is verified.