Research On Obstacle Avoidance And Path Planning Of Multi-robot Formation In Complex Environmen

With the continuous development of science and technology,the robot application has become more extensive.The single robot has great limitations when it performs complex tasks such as cooperative handling,encircle and area search.Because of its good load capacity,scalability,flexibility and other advantages,the multi-robot system has gradually become the research hotspot of scholars.The multi-robot operational condition is often complex and changeable.Therefore,the two key issue of multi-robot system collaborative operation are formation obstacle avoidance and path planning.During task execution,robots can quickly complete formation and flexibly change formation according to task requirements,so as to efficiently carry out path planning and avoid obstacles.The multiple robot formation can also better reduce the movement conflict between robots.Therefore,this paper mainly focus on obstacle avoidance and path planning of multi-robot formation in complex environment.The main research contents are as follows:(1)The experimental platform was built and the kinematics analysis was carried out.Firstly,the robot operating system and simulation environment are analyzed.Second,the software and hardware of the robot physical platform are designed,and the physical platform is built.Finally,the kinematic model of the robot is analyzed to establish the theoretical basis for the following dynamic window method and formation obstacle avoidance.(2)The path planning of ordering robot in complex environment is studied.First of all,a single robot can achieve real-time autonomous mapping.On the one hand,in order to solve the problem that the traditional A* algorithm has too many turning times,the evaluation function of the traditional A* algorithm is improved by setting the obstacle rate,reducing the turning times and improving the efficiency of path planning.On the other hand,in order to avoid the problem of A* algorithm skew through obstacle vertices,a safety threshold is set to improve the security of robot in the process of global path planning.In order to prevent the path planning from falling into the local optimal,this paper designs the azimuth deflection Angle on the basis of the traditional dynamic window method,introduces the global path,and integrates the improved A* algorithm.(3)On the basis of single robot path planning,the problem of multi-robot formation control in complex environment is studied,and the problem of multi-robot formation cooperative obstacle avoidance and path planning is further studied.Firstly,the traditional pilot-following method is improved by setting up virtual robots,and a double closed-loop PID formation controller is designed to realize formation control.Then,the adaptive formation obstacle avoidance based on the improved fusion algorithm is designed,and the obstacle avoidance is classified according to the different obstacles encountered by the formation.Finally,to solve the problem that multiple robots need to quickly recover formation after formation avoidance,a mode switching threshold is designed to make the formation better switch between formation mode and dynamic obstacle avoidance mode.(4)Verify the practicability of this research content.First,the Gazebo toolkit was used for simulation experiments.Then,the effectiveness and feasibility of obstacle avoidance and path planning technology of multi-robot formation in complex environment are verified by real vehicle experiments based on the built platform.