Wheeled bipedal robots (WBRS) combine the terrain adaptability potential of legged robots with the motion efficiency of wheeled robots, but the terrain adaptability is affected by the control system. Aiming at the defect that the traditional modeling ignores the dynamic changes in head angle and center of mass height, this paper proposes a method of integrated dynamic modeling and hierarchical control. The posture balance optimizes the system performance index through the linear quadratic regulator (LQR) to control the in-wheel motor, and the state feedback matrix is designed to suppress the tipping caused by external interference. At the same time, the changes in head angle and center of mass height are included in the balance control variables. The center of mass height changes are fed back through the proportional differential (PD) control and virtual model control (VMC) algorithm to control the joint motor. Simulation experiments are carried out on multiple platforms to verify that the proposed method effectively improves the control robustness of the traditional wheeled bipedal robot through geometric-dynamic coupling modeling and LQR-PD hybrid control, providing a new method of complex terrain adaptive control.
Xu et al. (Mon,) studied this question.