Research on Localization and Mapping for Lunar Rover Based on RBPF-SLAM

The capability of autonomous navigation is very important for lunar rover exploring in an unknown environment. In this paper, a method of simultaneous localization and mapping based on Rao-Blackwellized particle filters (RBPF-SLAM) is adopted to improve the precision of inertial positioning for the rover and to build a 2D grid map for the environment. Lunar rover´s motion model is built by combining Strapdown Inertial Navigation System (SINS) with a kind of odometry model, and the observation model of LiDAR is built using a likelihood field (LF) approach. Then the traditional RBPF-SLAM algorithm is improved.First, the most recent observation and the global map built before are considered in the proposal distribution; second, a grid-based incremental mapping method is presented. The results of simulation experiment show that the precision of localization by SINS is improved significantly using this method of RBPF-SLAM and an accurate and consistent 2D grid map is built successfully.