Development Of Multi-autonomous Mobile Robot System Based On Lidar Localization And Navigation

Multi-robot systems are more flexible than traditional automation systems,which can perform more complex and difficult tasks,so in recent years,multi-robot systems have become popular in robotics.There is a clearer solution to set up more autonomous mobile robot system for collaborative control,for the sake of the development of localization and navigation technology of mobile robots,and of the research results in networking technology and multi-agent theory of multi-robot system.In this dissertation,LIDAR based localization and navigation technology of multi-robot systems is studied.With this technology,researchers can use robots to explore unknown environment and get 2D map for navigation.Furthermore,based on this technology,researchers can expand functions for multi-robot systems,broaden their application scenario,and enable them to complete collaborative tasks in dynamic environments.The following work is done in this dissertation.First of all,hierarchical and modular ideas were used to develop two feature-rich,scalable and autonomous mobile robot platforms.The two mobile robots use Mecanum wheels as drive wheels to obtain omni-directional motion abilities,which makes the robots adapt to indoor narrow channels.The hardware and software architectures of these two robots have advantages in extensibility.Researchers can access different external devices and functional modules to achieve modular enhancements.Moreover,the two mobile robots have precise movement accuracy,which carries out a good foundation for the follow-up research work.Secondly,the simultaneous localization and mapping(SLAM)technology of mobile robot and the localization navigation technology in known environment are studied.The robots use LIDAR and the Monte Carlo Localization algorithm based on particle filter to obtain the global localization information.In unknown environments,LIDAR SLAM can solve the localization problem and can build maps.In known environments,in order to reach the destination,the robots use the A* path finding algorithm to plan global paths in the static global map,and use Trajectory Rollout and Dynamic Window approaches to plan the local motion path in the dynamic local map,so that the robots can follow the optimal route to travel automatically,avoiding obstacles at the same time,and can arrive at the destination in the end.After this,a number of experiments were designed to test the localization and navigation function of the robots.The motion accuracy,SLAM building function,autonomous navigation function and human-following function of the robots were tested respectively.On one hand,these experiments verified the excellent performance of the robot platforms,and verifies the feasibility of the robot localization and navigation algorithms in the test and their feasibility in practical applications.Then,based on the distributed networking features of ROS(Robot Operation System),this dissertation built a flexible and stable multi-robot system.This dissertation adopted a hybrid system architecture to carry on the cooperative control,taking the global optimization of system decision and the flexibility of sub-task coordination into account.At the level of movement coordination,the traffic rules method provided a simple and feasible idea of motion coordination.The coordination algorithm based on artificial potential field method provides a better coordination strategy in complex dynamic environment.In the end,a number of test experiments were designed to test the localization and navigation function of the robot system.The SLAM function,autonomous navigation function and human follower function were tested respectively.On the one hand,it verifies the positioning and navigation performance of the robot platform,and on the other hand,it verifies the feasibility of the localization and navigation algorithm in practical application.This dissertation designed a number of experiments for multi-robot systems.These experiments are in test environments and actual scenes to verify the comprehensive performance of multi-robot systems.For the homogeneous multi-mobile robot system,this dissertation used cooperative control methods to realize combined motions of two mobile robots,which verifies the effectiveness of the methods.For the heterogeneous multi-robot system,this dissertation designed a collaborative mission demonstration experiment in actual office scenes.The two mobile robots work together with an industrial robot arm formed a team to complete the document delivery and archiving task.This experiment tested the actual performance of the multi-robot system in a realistic scenario.