A Physical Parameter-Based Skidding Model for the Snakeboard

Abstract

The physical violation of a nonholonomic system's idealized constraints in the form of skidding has recently elucidated interest in new models in order to directly incorporate the phenomenon into the system dynamics. However, such models either are too simple to capture physical attributes or have otherwise been tested only on systems with simple behaviors, such as the rolling disk. In this work, we present a novel skidding model, based on physical parameters, for a snakeboard system, which is simultaneously rich in behavior but simple in design. This model extends the system's configuration space and associates traction forces to a skidding angle using an experimentally verified observation from the literature. We validate our model in simulation and discuss its advantages over the Rayleigh dissipation function skidding model. We also show that the model accurately predicts a physical system's behavior in experimentation with tuned parameters and standard controllers.