The dynamics of electric unicycles with human rider: modelling and experiments

In this work, a variety of experiments on different maneuvers of a human-ridden electric unicycle (EUC) are conducted. Position and orientation data of the human body segments are extracted from a motion capture system. We establish a skeleton model for the human (while riding the EUC) based on the collected experimental data. It is demonstrated that the fluctuations in the distance between the rider's center of gravity and the wheel center, and in the rider's mass moment of inertia are small during the maneuvers. We focus on the longitudinal dynamics and, by replacing the human with a hinged rod, we derive a three degree-of-freedom model, where the rider can lean independently from the EUC frame. Using the derived 3 DoF dynamical model, we identify the rider's mass properties from the experimental data and compare those parameters to the ones calculated from the skeleton model. A proportional–derivative control law is also identified for the EUC driving torque. The derived longitudinal model with data-driven parameters and the identified control law from the deceleration data provide good agreement on the simulations, even when testing them on different maneuvers. This demonstrates the predictive power of the developed modelling approach.