Adaptive robust control for a parallel orthopedic rehabilitation robot with guaranteed healing-safety

In orthopedic rehabilitation, ensuring both safety and precision of the mechanical stimulation is paramount for effective bone healing. This study proposes a safety-guaranteed control strategy for a novel 6-DOF orthopedic rehabilitation robot, which can deliver active multi-axis mechanical stimuli to tibial fractures. To address the stringent safety requirements of clinical settings, an adaptive robust control architecture was proposed to incorporate prescribed tracking performance. The core innovation lies in the integration of a piecewise time-varying performance function, which strictly confines the robot’s motion error within a predefined safe envelope during both dynamic transient and steady-state phases. To handle the system’s inherent nonlinearities and time-varying biological disturbances, a lightweight neural network is employed to adaptively compensate for lumped uncertainties in real-time, augmenting an integral sliding mode controller (ISMC). A series of hardware experiments on a self-developed 6-DOF orthopedic rehabilitation robot platform validate the effectiveness of the proposed method. The experimental results indicate that the controller effectively suppresses disturbances and maintains tracking accuracy within the predefined bounds, which demonstrates strong potential to enhance safety and tracking reliability in robotic orthopedic rehabilitation.