Position-Tracking Controller for Two-Wheeled Balancing Robot Applications Using Invariant Dynamic Surface

This paper suggests a position-tracking algorithm for the outer-loop through the invariant dynamic surface approach for balancing robot applications. The main feature is to devise a dynamic surface representing the target position-tracking performance with variable cut-off frequency. The proposed controller makes the dynamic surface invariant while updating the closed-loop cut-off frequency accordingly with the self-tuner. The closed-loop properties are rigorously analyzed. The experimental verification result shows that the proposed controller establishes the 48% enhancement of the circular tracking performance in comparison with the feedback linearization method, where the LEGO Mindstorms EV3 is used.