The Design And Implementation Of A Robot Navigation System Based On Lidar

With the rapid development of social services,warehouse logistics,space exploration,medical care,exploration and rescue industries,autonomous navigation mobile robot is facing a huge application requirement.More and more complex and changeable application environment puts forward higher requirements for mobile robot navigation technology.In this thesis,the basic principle of navigation function of mobile robot is deeply analyzed.Combining with ROS operating system,the navigation system of mobile robot based on lidar is designed to realize autonomous navigation.Firstly,in the process of building environment map for mobile robot,aiming at the inaccuracy of position and pose estimation of mobile robots caused by the fact that the latest observation information is not taken into account when constructing environmental map based on RBPF-SLAM algorithm,an improved Gmapping——SLAM algorithm is adopted to construct environmental map.Because the traditional artificial control method of mobile robot cannot meet some accurate target tracking location and mapping,this thesis proposes a method of map building based on Camshift target tracking to guide the movement of mobile robot to the sample point.Secondly,the basic principles of global path planning A_star algorithm and local path planning artificial potential field are analyzed in the path planning of mobile robots.Aiming at the problem of path planning failure caused by minimum point of Artificial Potential Field,this thesis proposes adding temporary virtual potential field at the point to remove the minimum area,so as to ensure the smooth navigation of mobile robots.At last,in the experimental part,this thesis verifies the validity and feasibility of building environment map for mobile robot based on Camshift target tracking and guidance and realizes the path planning and completes the navigation function of mobile robot.