Research On SLAM And Navigation Technology Of Intelligent Car Based On Lidar

With the continuous development of intelligent logistics vehicle technology and autonomous vehicle driving technology,various forms of intelligent vehicles have been widely used in industry,and more and more researchers have started to devote themselves to the research work in the field of intelligent vehicles.The exploration of unknown environment has been the hot spot and difficult point of its research,among which simultaneous localization and mapping(SLAM)and path planning scheme are the key technologies for intelligent vehicles to realize autonomous localization and navigation,and it has far-reaching guiding significance for the development of intelligent vehicles by studying its related algorithm technology.In this thesis,we mainly focus on the problem that intelligent vehicles cannot make full use of GPS information for localization and navigation in indoor environment,and study the LIDAR SLAM algorithm and path planning algorithm,and apply them to the intelligent trolley mobile platform to solve its localization and map building problem in indoor environment and finally realize autonomous navigation,the main research contents are as follows:Firstly,the overall framework for realizing the SLAM navigation work of the intelligent trolley is built,the relevant mathematical models are established,including the Li DAR model,the intelligent trolley motion model and the IMU model,and the construction process of the grid map model is analyzed and derived to prepare for the research of the SLAM and navigation algorithm of the intelligent trolley.Then,a theoretical analysis and derivation of the particle filtering algorithm is carried out,and the RBPF-SLAM algorithm is introduced to solve the map construction problem of the intelligent cart.To address the problems of low positioning accuracy and loss of particle diversity in the traditional RBPF-SLAM algorithm,the EKF is used to fuse the front-end wheel speed odometer and IMU information to improve the positional estimation accuracy,to remove the point cloud distortion of the Li DAR,and to improve the proposed distribution based on the fused motion model and observation model.algorithm particle scarcity problem.The map construction effect of the two SLAM algorithms is compared by using public datasets and Gazebo simulation experiments,and the superiority of the improved SLAM algorithm in terms of map construction accuracy effect is initially verified.Next,for the autonomous navigation work of the intelligent cart,a path planning method integrating the improved A* and TEB algorithms is proposed,and the traditional A* algorithm is improved by optimizing the heuristic function,introducing strategies such as safety expansion and key point extraction,and smoothing and optimizing the output trajectory,and the improved A* algorithm is integrated with the TEB algorithm as the path planning method to achieve the autonomous navigation task of the intelligent trolley,and the performance of the navigation algorithm is tested through simulation experiments to verify the feasibility and effectiveness of the navigation algorithm in this thesis.Finally,by deploying a distributed communication network for control and visual monitoring of the remote intelligent trolley Nanorcar,a functional package of SLAM algorithm and path planning algorithm is designed and integrated in the ROS system,and a real vehicle experiment of the intelligent trolley is conducted in a selected real indoor environment to verify the usability and superiority of the studied SLAM algorithm,and the navigation of the intelligent trolley to a specified target point is completed on this basis to verify the effectiveness of the proposed path planning algorithm.