文献类型：会议论文

英文题名：An improved LIDAR SLAM algorithm of self-driving

作者：Chenxi, Li[1]; Jun, Zhang[1]; XinYu, Jin[1]; GuangJing, Li[1]

第一作者：Chenxi, Li

通讯作者：Chenxi, Li

机构：[1] College of Robotics, Beijing Union University, Beijing, 100101, China

第一机构：北京联合大学机器人学院

会议论文集：IWACIII 2017 - 5th International Workshop on Advanced Computational Intelligence and Intelligent Informatics

会议日期：November 2, 2017 - November 5, 2017

会议地点：Fuxing Road #3, Haidian District, Beijing, China

语种：英文

外文关键词：Intelligent computing - Iterative methods - Optical radar - Particle swarm optimization (PSO)

摘要：In order to realize autonomous navigation of autonomous vehicles that don't rely on GPS, an improved SLAM method based on FastSLAM algorithm is proposed in this paper. First, a large number of data obtained by lidar are preprocessed, including: Remove sparse points, Voxel Grid, point cloud separation. Then, the processed data is used to make the point cloud match of the adjacent frame by using the method of PLICP(Point-to-Line Iterative Closest Point)algorithm to obtain the coarse positioning of the autonomous. Then the particle scatter points are carried out near the rough locations, and the niche technique particle swarm optimization method is introduced to locate the autonomous vehicle accurately. The map of lidar point cloud is represented by raster map. After getting the precise positioning of the autonomous vehicle, a local map which carried by the particle is added to the global map to update the map. Finally, the system outputs the map of the environment and the trajectory of the autonomous vehicle. The validity of the proposed method is proved by the experiment on autonomous vehicle platform.