Department of Systems Engineering and Automation DEPARTMENT OF SYSTEMS ENGINEERING AND AUTOMATION UNIVERSITY OF MÁLAGA Path Planning for Reconfigurable Rovers in Planetary Exploration Authors: J. Ricardo Sánchez 1 Carlos J. Pérez del Pulgar 1 Martin Azkarate 2 1 Universidad de Málaga 2 European Space Agency
25
Embed
Path Planning for Reconfigurable Rovers in Planetary ... · 1. Introduction • Objectives – Autonomy on rovers ← Path Planning – Implementation of Path Planning Algorithm for
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Dep
artm
ento
f Sys
tem
sEn
gine
erin
gan
d A
utom
atio
nDEPARTMENT OF SYSTEMS ENGINEERING AND AUTOMATION
UNIVERSITY OF MÁLAGA
Path Planning for Reconfigurable Roversin Planetary Exploration
• ExoMars mission– Place a rover on the Martian surface by 2021– Find signs of current or past life on Mars– Kinematic Configuration of Rover: 6x6x6 Degrees Of Freedom– Can provide an extra mode of locomotion: Wheel-walking
• Improves traction on loose soils
1. Introduction
• Objectives– Autonomy on rovers ← Path Planning– Implementation of Path Planning Algorithm for reconfigurable rovers
• Previous approaches:– E. Rohmer et al. Dynamic simulation-based action planner for a
– 6 Driving Joints• In charge of rolling wheels• Commanded in speed
– 4 Steering Joints• In charge of turning wheels• Commanded in position
– 6 Deployment Joints• In charge of deploying legs• Commanded in both position andspeed
• Virtual Model in V-REP:
2. Simulation Scene
• ExoTeR Locomotion ModesNormal driving• Walking Joints are locked• Maneuvers:
• Ackerman Turns• Spot Turns
Wheel-walking• Steering Joints are locked• Legs from one side drag the
rover while the others advance
2. Simulation Scene
• Martian Surface– Point cloud from Planetary Robotics Lab (PRL) at ESTEC– From point cloud to square grid with elevation data– Heightfield model in V-REP from grid– Texture from orthonormal image, modified to show two types of terrain
Loose Soil
Compact Soil
2. Simulation Scene
• Martian Surface– Each type of soil has different values of friction and slippage– Loose Terrain
• Best to use Wheel-walking– Compact Terrain
• Best to use Normal driving
TERRAIN INTERACTION
MODELING
2. Simulation Scene
• Cost Map– Detect Obstacles (high slopes and dangerous elements)– Make Obstacle Area Larger– Recognize soils– Add risk value