An Improved Hector SLAM Algorithm based on Information Fusion for Mobile Robot

In degeneration circumstance such as corridor, due to lacking environmental features, the traditional laser SLAM cannot locate the position of the mobile robot precisely during the scan matching, and further lead the mapping does not confirm to the reality. Aim at this problem, this paper proposes an improved Hector SLAM algorithm based on information fusion for mobile robot. Odometry can accurately reflect the distance of robot movement and Inertial Measurement Unit (IMU) can exactly show the angle of robot movement. So an improved SLAM algorithm framework is set up. In the process of information collection and handling, based on extending Kalman Filtering (EKF) fusing the information of odometry, IMU and laser scanning, to raise the accuracy of the change of point cloud to provide precise input information for scan matching process. In the scanning matching process, Hector SLAM algorithm is used to project the scan information onto grid map. In the last, according to the proposed method, the algorithm function packages are written under the Robot Operating System (ROS). The effectiveness of the method is verified by simulation experiment and robot platform physics experiments.