Terrain Slope Disturbance Compensation for Driving Performance of a Self-balancing Mobile Robot

When a self-balancing mobile robot moves on a ramp, the gravitational moment acting on the inverted pendulum body and two parallel wheels is greatly changed depending on the uncertain slope of inclined terrains, which can be regarded as a major environmental disturbance degrading the driving performance and the pitch balancing stability. This study investigates a real-time estimation algorithm for the unknown terrain slope in terms of the nonlinear disturbance observer (NDOB) technique. To this end, we establish a dynamic model of the self-balancing robot including the slope parameters of a three-dimensional terrain. Finally, we propose a sort of navigation-free terrain following control method to simultaneously compensate for the longitudinal as well as the lateral slope. The effectiveness of the proposed method is demonstrated through simulations and experiments.