Interaction dynamics modeling and adaptive impedance control of robotic exoskeleton for adolescent idiopathic scoliosis.

Adolescent Idiopathic Scoliosis (AIS), an abnormal lateral curvature of the patient's spine with vertebral rotation, is one of the biggest problems in the orthopedic profession. Non-surgical treatments of AIS, including bracing, has the following limitations: i) sensor-less design and lack of closed-loop control on the motion and forces exerted on the torso result in no adaption to the skeletal changes during the treatment, ii) lack of interaction control may also make the brace inconvenient and insecure for the patient. In this paper, a novel robotic exoskeleton equipped with linear actuators, position, and force sensors is developed to control the motion and forces exerted on the torso. It consists of three Stewart-Gough Platforms (SGP). Robotic AIS rehabilitation is modeled by combining the dynamics formulation of each SGP with the mass-spring-damper model of the thoracic segment and utilizing the inverse dynamics problem to compute the required forces for moving the spine to the desired position. Numerical ADAMS and MATLAB simulations are used to verify the interaction dynamics formulation. In addition, impedance control is used to control the interaction instead of using position and force control separately. A Model Reference Adaptive Impedance Control (MRAIC) is also proposed to compensate for the negative effects of uncertainties existing in the interaction dynamics modeling and improve the interaction control. The performance of the MRAIC is verified and validated using numerical MATLAB simulations and real-time experiments in terms of position tracking and reference impedance model tracking.