| Localization is the most basic problem of mobile robot autonomous navigation.In outdoor environments, sonar, laser range sensors and vision sensors are commonsensors which used to detect physical feature or landmarks to determine the pose ofthe robot. But these methods are vulnerable in some outdoor environment. Globalnavigation satellite, such as GPS, is popular method for outdoor robot localization,but satellite signals may be intentionally disturbed or distorted by trees or blocks.So it can mot meet the requirement of robot localization precision at any time and atany time.A new terrain inclination based localization technique is proposed in this paperto allow the robot to identify its three-dimensional location, which can be analternative method of localization when GPS is lost for a period of time and othersensors are not suitable to use. Given a topographical map or points cloud map anda planned path, a robot-terrain inclination model is used to abstract the relationshipof robot position and attitude along the path on the terrain that the robot is operatingon. This relationship is saved in a lookup table as the measurement function oflocalization. In order to solve the large uncertainty of measurement model, particlefilter is used to fuse the measurement data with the robot motion to achievelocalization.Three outdoor experiments were carried out to validate the proposed terrainlocalization technique. The first one was implemented at a small experimental sitewith a short path under different initial conditions, and the topological map wasobtained with an optical capture system. The second one was implemented at thesame experimental site using points cloud map. The third experiment wasimplemented at a larger and terrain-rich experimental site with a long path usingpoints cloud map. Experimental results show that the proposed terrain inclinationlocalization method could achieve satisfying localization performance on inclinedoutdoor terrains. |
PDF Full Text Request