Distributed Multi-robot Path Planning Based On Ant Colony Optimization

Robot path planning refers to plan a collision free optimal path from the starting point to the target point for the robot according to a set index in the environment with obstacles.Multi mobile robot path planning is one of the key problems in the field of logistics robot at present,which has important research value in manufacturing,warehousing,medical services and other fields.This paper mainly studies the distributed path planning strategy of multiple mobile robots,explores the application of bionic algorithm in mobile robot path planning,and designs a simple conflict resolution algorithm to make multiple mobile robots work efficiently in the shared workspace.This paper presents an improved ant colony optimization for robot path planning.The improvement of ant colony optimization mainly aims at solving the problems about strong randomness,easy deadlock and more iterations of classical ant colony optimization.On the basis of building an environment model with grid method,a kind of guidance function pointing to the target point is designed to reduce the blindness of ant search in the initial stage,so as to solve the problem of strong randomness.Furthermore,by increasing the threshold and path weight of pheromone update formula,the influence of excellent ants on path selection is increased,and the number of iterations is reduced.At the same time,in order to reduce the occurrence of deadlock,a virtual obstacle marking method is designed to actively avoid deadlock.Finally,the simulation results show that the improved ant colony optimization speeds up the convergence speed and has strong path search ability and adaptability.In order to improve the plug and play performance of the algorithm,a bi level programming strategy is adopted to design a distributed multi robot path planning algorithm framework.The first layer uses the improved ant colony optimization,and each mobile robot plans its best path locally.In the second level planning,the rolling window method is used for conflict detection.Each robot detects whether there are other partners in its adjacent area,and judges whether there is a collision by exchanging information.Then,based on the priority determination,a multi robot conflict resolution algorithm is designed to achieve local avoidance and solve the space-time conflict problem of the planned path.The experimental results show that the pause time of the multi robot system is the shortest and the path cost is small under the condition of keeping the longest path unchanged,which verifies the feasibility and superiority of the algorithm.