Research On Multi-robot Cooperative Task Allocation And Path Planning

With the advent of Made in China 2025 and industry 4.0,robot has great potential in the fields of intelligent factory,intelligent production and intelligent logistics.With the rapid development of key technologies such as artificial intelligence,operation research optimization,computer,control theory and communication,multi-robot system has attracted wide attention and becomes a research hotspot in related fields at home and abroad.Compared with a single robot,multi-robot system has the characteristics of distributed function,easy expansion,strong fault-tolerance,and has great advantages in dealing with complex and tedious tasks.However,multi-robot system is also facing many difficulties and challenges.Task planning which plays a vital role in the robots cooperation is one of the difficult problems.In this paper,the task planning of multi-robot system is divided into task allocation and path planning.The development history and research status of multi-robot system,task allocation and path planning are discussed.(1)In order to solve the multi-robot task allocation problem considering execution time sequence,an improved ant colony optimization algorithm is proposed.This problem can be considered as a multi-traveling salesman problem model.Firstly,ant colony optimization algorithm is used to optimize the solution space.Then mutation operator in genetic algorithm is introduced to optimize the task order of each robot.Finally,Metropolis criterion in simulated annealing algorithm is introduced to jump out of the local optimum.(2)In order to solve the task allocation problem of multi-robot formation,a bi-level programming model is proposed.According to the existing robot formation cooperation mode,an upper-level model of robot formation is established.The improved discrete particle swarm optimization algorithm is used to obtain the number of robot and best formation of each task.According to the existing formation library,a lower-level model of the robot location is established,and the robot position in the formation is obtained by using the genetic algorithm.(3)In order to solve the problem of multi-robot cooperative path planning and path replanning,considering the constraints such as collision avoidance,environmental obstacles and so on,and taking the length of the planned path and the degree of smoothness of the path as performance indicators,an improved rapidly-exploring randomized trees(RRT)algorithm is proposed to plan a safe optimal path for every robot from the starting point to the target point.On this basis,the above algorithms are analyzed and verified by simulation experiments.The results show that the designed algorithms can effectively solve the problem of multi-robot task planning.