Research On Task Assignment And Path Planning For Multi-Robots For Intelligent Warehouse

Mankind can achieve the purpose of reducing labor intensity and improving efficiency by using multi-mobile robot system to perform monotonous,complicated and even dangerous tasks in fields such as industrial and agricultural production.In the field of storage and logistics,intelligent warehouse has gradually become the inevitable trend of the development of productivity,using multi-robot system to achieve warehouse automation and intelligent is a viable method.This paper is mainly based on intelligent warehouse system for multi-robot task assignment and path planning research,and builds a multi-robot system for intelligent warehouse for testing.Task allocation is the basis of making full use of system resources and improving system efficiency,its importance increases with the functional differences of system members and the structural complexity of the task.According to the characteristics of multi-robot system such as environment variability,task diversity and dynamism,this paper presents an environment description model,task model and robot model for multirobot system under warehouse environment,analyzes the classical integer planning method,graph search method and so on,a multi-robot task assignment algorithm based on market mechanism is proposed.Under certain optimization criteria,the path planning problem is to find an optimal path in the environmental map to complete the tasks.The characteristics of multi-robot path planning are analyzed,and the storage environment map model and the robot collision avoidance strategy are proposed.The classical global and local robot path planning algorithms are studied.Because the traditional A * algorithm does not consider the time dimension,leading to collision and excessive waiting of the robot path planning,an A* algorithm with time domain collaboration is proposed.It designs robot positioning method and motion control method,builds a multi-robot test platform for intelligent warehousing.Based on this platform,the robot's positioning and motion control accuracy are analyzed.In the method of task allocation based on market mechanism,the performance of two indicators of the total distance and the longest time to complete the task are compared,and when the two indicators are fused under a certain proportion,the cost of task is lower.The path planning of multi-robots with time dimension A* is analyzed,and then path conflicts among multiple robots are avoided in advance.So it solves the problem of path planning of multi-robots system well.