The Research On Task Scheduling And Path Planning For Mobile Robot Cluster

Mobile robot technology has been widely used in intelligent manufacturing,medical care,patrol monitoring,rescue search and other fields.Compared with the single mobile robot system,the mobile robot cluster system has good efficiency,parallelism,scalability,and robustness.The task scheduling technology and path planning technology of mobile robot clusters are two important and difficult issues in mobile robot cluster systems.This paper conducts in-depth theoretical research and practical exploration on task scheduling and path planning of mobile robot clusters.In this paper,an improved genetic algorithm is used for task scheduling of mobile robot clusters.First,we use double chromosomes to encode individuals.One chromosome contains all the tasks that all robots need to complete,and the other chromosome contains the breakpoint index that cuts the first chromosome.By decoding the double chromosomes,the task sequence of each robot can be obtained.Then,the randomly obtained initial population is divided into several small populations,and the eight genetic operators are used to perform genetic operations on the small populations in parallel.Finally,the fitness of the individual is obtained through the fitness function.The roulette algorithm is used to select individuals with high fitness in the population for inheritance.The population optimality is finally achieved through population iteration,and the best robot task scheduling scheme is obtained.Experimental results show that,compared with the traditional genetic algorithm,the improved genetic algorithm is suitable for both single mobile robot task scheduling and mobile robot cluster task scheduling.It has the advantages of low complexity,fast convergence,and low system overhead.This paper compares the path planning experiment of mobile robot using Dijkstra algorithm,best priority search algorithm,and A * algorithm in obstacle environment.The experiment shows that A * algorithm as a heuristic directed search algorithm has better path search performance.Therefore,this paper chooses A * algorithm as the path planning algorithm of mobile robot.In addition,in actual situations,considering the physical size of the robot,the robot is likely to collide with obstacles at corners.Therefore,this paper proposes an improved A * algorithm that considers the physical size of the robot to avoid the collision of the robot with the obstacle at the corner.In this paper,a large number of path planning experiments are carried out in the warehouse environment and random obstacle environment.The experiments show that the improved A * algorithm can get an efficient and collision-free optimal path.This paper verifies the effectiveness of mobile robot task scheduling algorithm and path planning algorithm in actual scenarios.First of all,the intelligent warehouse simulation environment was built using the MATLAB platform.The mobile robot cluster carried out cargo handling in the intelligent warehouse.The task scheduling algorithm,path planning algorithm and path conflict resolution were used to ensure the normal operation of the intelligent warehouse.The simulation shows that in a fixed-size warehouse,a small number of robots can efficiently complete multiple cargo sorting tasks in parallel.Too many robots will cause the system complexity to be too high,and the robot cluster system will deadlock.Then,build a physical robot embedded with the ROS robot operating system,and use the physical robot to complete the indoor map construction task,target search task and multi-point cruise task.