Robust transition control of underactuated two-wheeled self-balancing vehicle with semi-online dynamic trajectory planning

Abstract In order to increase the mobility and dynamic agility of a two-wheeled self-balancing vehicle, a dynamic posture control strategy for the initial state transition is proposed in terms of a well-planned nonlinear reference trajectory. The dynamic trajectory planning for the underactuatedmobile inverted pendulum is conducted by solving a two-point boundary value problem with the constraint equation of internal dynamics. The restriction of the offline trajectories for real-time applications is relaxed by establishing a semi-online planning method by assuming velocity-controlled missions. For the transition tasks requiring high maneuver motions including quick start and stop and quick speed change, the proposed control scheme is profitable to stabilize the initial transient motion and get over the performance limit of the conventional feedback-only control with typical step reference inputs. As a result, it enables to achieve consistent state transition to the target velocity and requires just minimal control efforts.