Application Research On Scheduling And Coordination Of Multi-Handling Robot Based On Colored Traveling Salesman Problem

In order to cope with the rapid growth of goods sorting operations in e-commerce warehouses brought about by the booming e-commerce,major e-commerce companies have deployed a large number of sorting robots to complete the overall handling of the shelves.Task scheduling is mainly based on the principle of closest.Although it is simple and easy,it affects the full play of multi-robot system capabilities,and there is a lot of space for optimizing scheduling.The task scheduling and coordination problem of the multi-handling robot system itself is computationally complex and NP-hard.This limits the construction scale of the sorting system.At present,the maximum size of a single system is limited to about 300 units.In addition,at present,foreign countries have successfully developed picking robots that pick up goods one by one,but key technologies such as cluster scheduling,path planning,and coordinated control are waiting for research and exploration.Sorting and picking scenarios are urgently needed to consolidate the technical basis of task scheduling and path planning for multi-robot handling systems,and to develop efficient modeling techniques and solving algorithms.In this regard,this paper takes the Colored Traveling Salesman Problem(CTSP)theory as the main basis,and uniformly studies the multi-handling robot task scheduling,path planning and coordinated control methods in the two scenarios of sorting and picking.The specific research contents and results are as follows:Firstly,the CTSP modeling and solving of the task scheduling problem of multi-handling robots is studied.The task scheduling problem of multi-sorting robot and multi-selecting robot is modeled into two types of CTSP respectively.The genetic algorithm(GA),firefly algorithm(FA)and variable neighborhood search algorithm(VNS)for solving CTSP are given.The grouping parallel solution strategy is proposed to realize the task scheduling of large-scale handling robots in a short time.The results of a lot of simulation experiments show that under the two scenarios of sorting and picking,the scheduling result of VNS is much better than the principle of proximity.VNS realizes optimal scheduling of 500 handling robots and 50,000 tasks within 10 s,which exceeds the existing maximum single schedule size.Then,the path planning problem of the handling robot in the e-commerce warehouse is studied.This paper introduces the e-commerce warehouse environment and uses the grid map method to establish the grid electronic map of the e-commerce warehouse,and implements the path planning method based on Dijkstra algorithm and improved A* algorithm.Simulation experiments show that the improved A* algorithm outperforms the Dijkstra algorithm in path optimization and time consumption,and can be used preferentially.Finally,the coordination problem of multi-handling robot system is studied,and the multi-handling robot scheduling and coordination simulation software is developed.In this paper,the various path conflicts of multi-handling robots are merged into three types: the relative conflict,the intersecting conflict,and the chasing conflict between the two robots.By giving the vehicle three different operations,different priority,delay waiting,and path re-planning,and the path conflict resolution strategy is established.The causes of blocking and deadlock of multi-handling robots are analyzed.The path occupation and abandonment strategy of multi-handling robots based on priority is proposed to realize the gradual elimination of deadlock and blocking.Finally,a multi-handling robot system coordination algorithm that integrates path conflicts,deadlocks,and abnormal conditions is formed.Using the developed simulation software,the effectiveness of the proposed CTSP-based scheduling scheme,path planning algorithm and coordination scheme is tested and the overall performance of the scheme is demonstrated.