Path Planning and Execution for Planetary Exploration Rovers Based on 3D Mapping
Andres Mora, Keiji Nagatani, Kazuya Yoshi · InTech eBooks · 2010
In this chapter, the combination of two path planning strategies, a global path planner and a local path planner, was discussed for a mobile robot (rover) to explorer in a lunar surface environment with poor illumination and rich obstacles. The global path planner would use the representation of the elevation data called the Global DEM (digital elevation map) gathered by a satellite mission orbiting the targeted celestial body to calculate a general path to the desired destination. The local path planner in the other hand, uses the representation of the elevation data called Local DEM obtained by its laser range sensor mounted on the rover to compute a path that would reconfigure the global path initially determined. An analysis of the LDEM, or local map, was made to classify the different areas according to the occlusion effect due to the presence of obstacles. This analysis allowed the quantification of the occluded and non-occluded regions within the area that can be covered by a laser range sensor. The algorithm that quantifies the non-occluded areas that can be perceived by the sensory system from the position where the rover is located within a local map was addressed. This quantification is given in the form of an index called the visibility index which is later incorporated in the local path planner through a method called the next sensing position. The next sensing position concept and its algorithm were explained thoroughly, and numerical simulation results were presented. The local path planner evaluates the terrain and finds an optimal path in terms of distance to the local goal obtained by the next sensing position algorithm.