Whole-Body Balancing and Differential Driving Control using a Relative Vector-Based LQR to Address Non-Holonomy of a Wheel-Legged Robot

Abstract

In this paper, we present a novel approach for controlling the balance of a wheeled bipedal robot (WBR) by utilizing the robot's dynamics and constraints. Our method is based on a hierarchical optimization framework, where we first use a relative vector-based linear quadratic regulator (LQR) to overcome the non-holonomic nature of the wheeled robot system and synthesize sagittal and lateral motion for a given reference trajectory. To validate the effectiveness of our proposed method, we conduct several simulations and the results demonstrate that it achieves superior performance.