Design And Implementation Of Visual Navigation System Based On ROS Wheeled Robot

Mobile robot is a research hotspot in the field of robotics at home and abroad.The research results of related theories and applications greatly improve the quality of current industrial production and social services.As the key technology of mobile robot research,autonomous navigation is the basis of intelligent and practical mobile robot.In view of the shortcomings such as poor universality and low navigation accuracy of current mobile robots,this thesis designs the visual navigation system for wheeled robot based on ROS(Robot Operating System)and realizes the functions of map construction and path planning.In order to solve the problems of poor code reusability and high coupling among functions in the process of robot development,this thesis analyzes the technical characteristics and operating mechanism of ROS in-depth.Combined with the functional requirements of autonomous navigation robots and the characteristics of ROS,the hardware structure and software framework of the visual navigation system for wheel robots are designed with the goal of low cost and high universality.In the construction of environment map,this thesis analyzes and compares the principles and key steps of RBPF-SLAM algorithm and Gmapping algorithm.Then,the motion control and sensor observation models of the wheeled robot navigation system are established.And the realization of the mobile robot map building function in the ROS is analyzed and parameterized.In the path planning,this thesis analyzes the principles and key steps of algorithms such as Dijkstra,A*,RRT,artificial potential field approach and dynamic window approach.In order to solve the problem of slow convergence and poor stability of the basic RRT algorithm in global path planning,an improved RRT algorithm is proposed by introducing the idea of variable probability target bias and the optimization strategy of node selection in random tree extension,which enhances the rapidity and stability of the algorithm.In the simulation and experimental parts,the comparison of the navigation effects verifies the feasibility and effectiveness of the Gmapping algorithm and the improved RRT algorithm.The flexible applications of ROS packages(stacks)and nodes in the experiments also validate the performance portability and extensibility of the navigation system.