Research On Path Planning Algorithm For Multi-Mobile Robot Systems

In recent years,with the rapid development of robot related technologies,multimobile robot systems have been widely used in various fields.The multi-robot path planning related algorithm is the key to ensure the successful application of multi-robot system.Therefore,it is very necessary and meaningful to study the multi-robot path planning related algorithms.Multi-robot path planning related algorithms mainly include multi-robot multi-task assignment,multi-robot global path planning,multi-robot real-time operation and coordinated obstacle avoidance.Among them,the multi-robot global path planning algorithm is the focus of multi-robot path planning related algorithms.This paper has made some in-depth research on these algorithms included in the multi-robot path planning related algorithm.Firstly,this paper establishes a mathematical model for multi-robot multi-task assignment problem and solves the problem by using improved genetic algorithm.In this paper,the mutation decision condition of genetic algorithm is improved.The algorithm retains the optimal individual to the maximum extent during the iterative process and thus increases the convergence speed.And combined with the multi-robot multi-tasking mathematical model,a new way of crossover is adopted.Secondly,the establishment of several multi-robot operating environments is studied.The grid method is used to establish the map environment based on the actual conditions.Then,the A* algorithm is used as the basic algorithm for multi-robot global path planning by comparing the common path planning algorithms.The focus of this paper is proposing a set of multirobot traffic rules suitable for multi-robot cooperative operation for several types of conflicts between multiple robots.And by introducing the vehicle traffic flow theory,the advantages of the multi-robot traffic model under the traffic rules of the paper are demonstrated.And by establishing a map information table and combining traffic rules,an improved A* algorithm is proposed for global path planning of multi-robots.The algorithm improves the efficiency of multi-robot global path planning by adding weight items to the A* algorithm evaluation function.This paper also proposes a multi-robot collaborative obstacle avoidance algorithm based on the traffic rules and the idea of BUG algorithm.And successfully combined it with the global path planning algorithm of this paper.Finally,the paper designs and implements a set of simulation software for multirobot completing multi-task in intelligent storage environment by combining the proposed theory and improved algorithm of this paper.The simulation system adopts a combination of centralized control and distributed control.The simulation results prove the applicability and high efficiency of multi-robot multi-task assignment based on improved genetic algorithm,global path planning algorithm based on traffic rules and improved A* algorithm,and real-time obstacle avoidance algorithm of the paper.