Research And Application Of The Control System Of The Ground Low-speed Robot Cooperative Formation

With the development of technology,single robots can no longer meet people's needs in many aspects because of its limitations,the research of robot is developing towards multi-robot system,but the collaborative control of multi-robot formation has always been a difficult problem.This paper takes the low speed robot on the ground as the research object,and studies the collaborative formation problem of multiple robots in an indoor environment.The main research contents include:(1)The omnidirectional mobile robot based on Mecanum wheels is built as the experimental platform.The control system adopts the distributed control architecture which contains host computer and slave computer.The single-chip and motor drive module in the slave computer system were designed and the hardware of the host computer system was selected.The robot is equipped with an industrial computer and a lidar,and runs Cartographer algorithm to construct the indoor map.It uses the adaptive Monte Carlo positioning algorithm to achieve autonomous positioning,which served as the basis for subsequent path planning and coordinated formation.(2)The fitness function is optimized and the elitist retention strategy is added to solve the problem of the unsmooth path and slow convergence speed in the genetic algorithm planning.The simulation results show that the improved genetic algorithm has faster convergence speed and smoother path,which is suitable for robot motion,and it is used in the global path planning;In order to avoid the random dynamic obstacles in the navigation process,the Dynamic Window Approach is adopted,and the effectiveness of Dynamic Window Approach is validated by simulation.(3)Aiming at the problem of obstacle avoidance of the leader-follower,the algorithm is improved by adding a virtual leader.According to the position and pose of the virtual leader and the set formation,the leader calculates the position and pose of the virtual leader in real time and sends it to the followers,the virtual leader guides the follower to move.The virtual leader guides the followers to move.PID controller is designed to help followers to track the pose of the virtual leader.In order to avoid dynamic obstacles that may be encountered during the formation process,a multi-robot collaborative obstacle avoidance strategy is designed The effectiveness of the improved leader-follower method is verified by simulation.(4)Experiment in a real environment.First,conduct global path planning experiments and local obstacle avoidance experiments,verifying that a single robot can use the improved genetic algorithm to plan the global path to reach the target in real environment,and can avoid dynamic obstacles;secondly,the cooperative formation experiment of multi robots in the barrier free environment is carried out to verify the effectiveness of PID controller;Finally,obstacle avoidance experiments in obstacle environments were conducted to verify the effectiveness of the improved leader-follower and obstacle avoidance strategies.