Open Access

Daniel Heidle, David GabrielORCiD, Gernot SchullerusORCiD, Georg SchmidtORCiD, Martin NeuburgerORCiD

• In this paper, three advanced control algorithms for riderless bicycle stabilization and yaw rate tracking are investigated and implemented on an experimental hardware setup. The control designs, attributed to linear quadratic integral (LQI) control, model predictive control (MPC) and optimal integral sliding mode control (OISMC), are based on the canonical linearized equations of motion from a Whipple bicycle model. Simulations evaluate each controller's robustness against model uncertainties and disturbances. The results of a comparative analysis demonstrate the performance of each control strategy, providing valuable insights for designing effective control systems that enhance bicycle stability and maneuverability in real-world scenarios. In addition, the effectiveness of adding a disturbance observer to the lateral dynamics controller is illustrated. Overall, the designed OISMC algorithm with disturbance observer shows superior robustness against model uncertainties and disturbances and is therefore recommended for stabilization and yaw rate tracking of self-riding bicycles.