| In recent years,indoor mobile robot technology is a hot direction in the field of robot research,and autonomous navigation technology is a key technology in indoor mobile robot technology.The research on autonomous navigation technology mainly includes three parts:autonomous mapping of mobile robots in unknown environments,detection technology for dynamic obstacles and path planning.This paper studies the autonomous navigation technology of mobile robots based on the research background of indoor mobile robots.The main research contents of this paper are as follows:(1)The autonomous mapping technology of indoor mobile robots for unknown environments is studied,and the RRT search algorithm is selected as the basis.The autonomous search strategy is a map construction algorithm based on the Gmapping-SLAM algorithm.The basic principle of traditional RRT algorithm for searching frontier points in complex indoor environment is studied,and frontier point optimization and its scoring mechanism are introduced to improve the efficiency of autonomous search.When the unknown area is beyond the search range of the mobile robot,the no frontier point coping mechanism is adopted to solve the dilemma of the mobile robot falling into no frontier point.The filtering-based Gmapping-SLAM map construction algorithm is analyzed.Aiming at the inaccurate point cloud matching and pose drift caused by the motion of the mobile robot to map construction,the lidar data combined with the odometry model is used to deal with the motion distortion.The mobile robot positioning technology based on Monte Carlo positioning is introduced in detail,and the methods of reducing resampling and limiting the number of particles are adopted to improve the mapping efficiency for the problems of particle dissipation and excessive particle information in the process of mapping.(2)The detection technology of indoor mobile robots for dynamic obstacles is analyzed,and the technical route of dynamic obstacle detection based on lidar sensor is determined.In-depth research on clustering segmentation technology is carried out,and a clustering segmentation algorithm combining distance threshold and density is adopted to improve the accuracy of clustering segmentation.The data correlation matching and target tracking technology between multiple frames are studied,so as to achieve a better detection and tracking effect of dynamic obstacles.(3)The path planning and obstacle avoidance algorithm of indoor mobile robot are analyzed,and the global path planning based on D*Lite algorithm and the local path planning based on artificial potential field method,path and speed planning are determined.After studying the basic principle of the pre-planning link of the D*Lite algorithm,the node screening mechanism is adopted to remove the redundant nodes brought by the search method.The method of D*Lite and artificial potential field method is combined to avoid the problem of being too close to static obstacles in the re-planning process,and for the local extreme value problem existing in the artificial potential field method,the local extreme value optimization strategy is adopted.to help the mobile robot out of trouble.The method of path and speed planning is used to avoid the dangerous collision between the mobile robot and the dynamic obstacles,so that the mobile robot has a good ability of path optimization and avoidance of dynamic and static obstacles.(4)The autonomous navigation algorithm designed in this paper is verified on the mobile robot platform equipped with 2D Li DAR and Mini PC.The results show that the mobile robot transplanted with the algorithm in this paper can realize autonomous mapping and dynamic obstacle detection technology.On the basis of,it can carry out safe and reliable path planning and avoid dynamic and static obstacles,which is feasible in actual scenarios. |
