Humanoid adaptive sliding mode control based on gait data

To address the requirements of lower limb rehabilitation robots for gait trajectory tracking accuracy and output torque smoothness, as well as the need for adaptive identification of the patient’s physical parameters. This study proposes a humanoid adaptive sliding mode control approach based on gait data. First, the gait trajectories of healthy individuals are collected using the NOKOV optical motion capture system. Second, the dynamic modeling of the lower limb rehabilitation robot is established through mechanism analysis. Then, a gait feature compensation and inertia compensation method are designed, and the model parameters are adaptively identified to enhance the generalization capability of the lower limb rehabilitation robot for different patients. Meanwhile, sliding mode control is introduced to improve the system’s disturbance rejection, and an interference prediction method is employed for compensation to suppress the chattering effect of sliding mode control. The stability of the control system is analyzed using Lyapunov’s criteria. Finally, the gait trajectories collected from healthy individuals are imported into the simulation system for real gait verification. The experimental results demonstrate that the proposed control method achieves rapid convergence of the trajectory tracking error while effectively suppressing high-frequency and high-amplitude jitter in output torque.