Terrain Feature Based Localization Mapping And Traversability For Mobile Robots

Mobile robots have been increasingly used in many outdoor applications such as planetary exploration, search and rescue, mine working where the robots are required to travel on uneven terrains. Localization and mapping are important problems for the mobile robot in outdoor complex environments. Hence this dissertation aims to slove three-dimensional localization and mapping problems of the outdoor mobile robot using the minimalist approach. The main contributions of this research work are summarized as follows:A new terrain inclination based localization technique is proposed to allow the robot to identify its three-dimensional location only using the onboard sensors in the outdoor environments when GPS is low or lost for a period of time. The assumption is that terrain characteristic map, such as a topographical map, can be obtained before the robots travel over the uneven terrain. Given a topographical map and a planned path, a Robot-Terrain Inclination model (RTI model) is extracted along the path on the terrain that the robot is operating on. A particle filter is then used to fuse the measurement data with the robot motion based on the extracted RTI model and achieve the localization.When the robot moves on the unknown environment that the terrain characteristic map can not be obtained beforehand, a new3D localization and mapping technique is proposed to allow a robot to identify its location and build a global map in an outdoor environment. The Iterative Closest Points (ICP) algorithm and terrain inclination based localization are combined together to achieve accurate and fast localization and mapping. Inclinations of the terrains the robot is navigating on are used to achieve local localization during the interval between two laser scans. Using the result of the above localization as the initial condition, the ICP algorithm is then applied to align the overlapped laser scan maps to update the overhanging obstacles for building a global map of the surrounding area. As the robot moves forward, the entire work space will be mapped finally.In order to provide more comprehensive environmental information for mobile robot in path planning, a new traversability assessment method is proposed to allow the robot to assess the traversability of the terrain online. Based on the onboard sensors, such as the wheel encoders and IMU, an extended Kalman filter (EKF) is used to estimate the wheel slip and tire forces for the mobile robot. The parameter which represents the characteristic of terrain type, such as the maximun traction force, will be extracted from the force-slip curve. According to the traction force and the slope of the terrain, a Bayes classifier based traversability assessment method is designed to assess how hard the robot could traverse on the new terrain.Comprehensive experiments were carried out to verify the effectiveness of several methods proposed above in a uniform framework. The experimental results show that the proposed methods could achieve good localization, mapping performance and online estimation of the terrain traversability. This research would be valuable addition to many autonomous robotic applications.