Research On Path Planning Of Multi Mobile Robots Unmanned Warehouse

Multi-mobile robot system is a common handling system in unmanned warehouse.The multi-mobile robot system with good handling efficiency can greatly accelerate the access efficiency of unmanned warehouse.Multimobile robot path planning technology is a key issue in multi-mobile robot system.The goal is to plan a collision-free path for each mobile robot in unmanned warehouse from the starting point to the end point,while ensuring shorter planning time and path length,so as to solve the "how to go" problem of multi-mobile robots.At present,the multi-mobile robot system has been applied in traditional factories,but the mobile robot in traditional factories moves in a fixed trajectory at a predetermined speed,which can meet the rhythm of the pipeline.However,the fixed trajectory and the preset speed could not fit for the randomness of tasks in unmanned warehouse and the distribution of task positions,so the path planning of multi-mobile robots in unmanned warehouse has become a research hotspot in recent years.This paper mainly studies the path planning method of multi-mobile robots for the handling task of multi-mobile robots in unmanned warehouse.It mainly includes the following aspects:1)Aiming at the problem that the traditional centralized multi-mobile robot path planning algorithm has a huge amount of computation in the highprecision planning map,a method combining rough-fine search strategy is proposed.Experiments show that compared with traditional algorithms,our algorithm has better performance in high-precision planning map.With the improvement of planning map accuracy,it maintains high and stable planning success rate,shorter planning time and better planning path.2)Aiming at the problem that the traditional distributed multi-mobile robots path planning algorithm is easy to fall into deadlock.Firstly,the solvability of the path planning problem of "lifelong" multi-mobile robots in strongly connected digraph is proved.Secondly,a distributed multimobile robots path planning algorithm based on reserved area is proposed,which can completely solve the problem.It could solve "deadlock" and collision problems in the working scenes of at least 20 mobile robots,and can still works well under the interference of dynamic obstacles.3)Aiming at the problem that the traditional multi-mobile robots path planning algorithm could not take advantage of the structural environment of unmanned warehouse.Firstly,the structural environment of unmanned warehouse is modeled,and the whole environment is abstracted into “workstation sub-model” and “channel sub-model”.Secondly,a mixed path planning algorithm for multi-mobile robots is proposed,which can solve “deadlock” and collision problems in the working scene of at least 8 mobile robots,and can still works well under the interference of dynamic obstacles.4)The simulation experiment platform and the actual environment experiment platform of the multi-mobile robot system are built respectively,and our algorithm is tested.The information of planning success rate,planning time and planning path length are tested.The experimental results show that our proposed algorithm is effective in multi-mobile robot path planning,and it has certain validity and practicability.