Control of Two-Wheeled Robots in Low-Traction Environments

Much research on two-wheeled robots has been completed in the past decade, but robot stability in low-traction environments has not yet been considered. A model for the two-wheeled robot including wheel slip effects is derived. A wheel slip dependent friction model is used to simulate two low-traction surfaces: the first has characteristics similar to ice and the second is frictionless, in addition to a no-slip environment. Two reference-tracking controllers designed for low-traction surfaces are described, and their effect on robot stability is compared to a typical controller designed for no-slip surfaces. The control objective of robot stability is tested by observing the tilt angle response of the robot, starting at rest with an initial tilt angle of 10◦. In simulations only the two new controllers are able to maintain robot stability on frictionless surfaces, and both outperform the baseline controller on other surfaces. A controller is used experimentally to achieve robot balance in the presence of mass imbalances and tilt angle measurement errors, using a reaction wheel actuator on a frictionless surface, but the controller is unable to balance on no-slip surfaces due to unmodelled tire effects.