We describe a novel six-wheeled "omnidirectional" autonomous robot platform based on the USU smart wheel mobility capability concept and a distributed, multiprocessor vetronics system. We also describe the multiresolution behavior-generation strategy that was developed for these systems. A task decomposition approach using a grammar of primitive maneuvers is the basis for mission and path planning for the ODV robots, and a feedback linearization control strategy is used for path tracking. Simulations and experiments show the effectiveness of the mission planning and control strategies. Our ongoing and future research considers the mobility advantages offered by the ODV platform over more conventional systems and looks for ways to more tightly integrate the mission and path planner functions with the vehicle-level control systems to develop intelligent, reactive behaviors.
Published in:
Control Systems, IEEE
(Volume:20
,
Issue:
6
)
Date of Publication: Dec 2000