| Self-localization is one of key technique for robotics.In this research,a single-line laser scanner is used as the main environmental perception sensor suite.The laser scanner combined with IMU(Inertial Measurement Unit)is utilized to build a self-localization and 3D mapping device.The main work is as follows:Firstly,the coordinate frame system and pose transformation approaches for robot localization and mapping are introduced.The coordinate system mainly includes the world frame,robot frame and sensor frame(e.g.Laser scanner).The robot orientation transformation mainly includes rotation matrix,Euler angles and quaternions.Secondly,we explain the basic principles of the IMU integral motion model.The relative motions between the two instant are derived bsaed on the IMU motion model.The IMU relative motion information is utilized to correct the distortion of lidar point cloud and to solve the environmental degradation problem.Then,a laser scanner is used to develop the real-time localization and 3D mapping technique.It includes four main parts:a single-line laser scanner spinning system integration;laser point cloud distortion correction;local laser odometer and global laser odometer.The spinning system enables the laser to acquire three-dimensional spatial measurement information.The local odometer can calculate the relative motion through the feature point cloud with high rate.The low-rate global odometer first corrects the point cloud of the local odometer and matches the corrected point cloud with a built global map.Finally,the global odometer and the local odometer are coupled to obtain high rate and high precision poses.This followed by an analysis and effective solution for environmental degradation problem.Finally,the performance of proposed system is investigated and analyzed based on the ROS(Robot Operating System)platform.The experimental results show the effectiveness of the developed system on robot self-localization&3D mapping. |
PDF Full Text Request