Neural-network-based nonlinear model predictive tracking control of a pneumatic muscle actuator-driven exoskeleton

• Published in: IEEE/CAA journal of automatica sinica Vol. 7; no. 6; pp. 1478 - 1488
• Main Authors: Cao, Yu, Huang, Jian
• Format: Journal Article
• Language: English
• Published: Piscataway Chinese Association of Automation (CAA) 01.11.2020; The Institute of Electrical and Electronics Engineers, Inc. (IEEE); Key Laboratory of Image Processing and Intelligent Control, School of Artificial Intelligence and Automation, Huazhong University of Science and Technology, Wuhan 430074, China
• Subjects: Actuation; Actuators; Algorithms; Approximation; Control stability; Control systems design; Controllers; Exoskeletons; Feedback control; Gait; Gaussian process; Gaussian processes; Legged locomotion; Mathematical analysis; Muscles; Neural networks; Nonlinear control; Optimization; Orthoses; Predictive control; Predictive models; Rehabilitation; Spinal cord injuries; Subsystems; System dynamics; Task complexity; Tracking control; Training; model predictive control; Echo state Gaussian process; neural network; pneumatic muscle actuators-driven exoskeleton
• ISSN: 2329-9266; 2329-9274
• DOI: 10.1109/JAS.2020.1003351

Pneumatic muscle actuators &#x0028 PMAs &#x0029 are compliant and suitable for robotic devices that have been shown to be effective in assisting patients with neurologic injuries, such as strokes, spinal cord injuries, etc., to accomplish rehabilitation tasks. However, because PMAs have nonlinearities, hysteresis, and uncertainties, etc., complex mechanisms are rarely involved in the study of PMA-driven robotic systems. In this paper, we use nonlinear model predictive control &#x0028 NMPC &#x0029 and an extension of the echo state network called an echo state Gaussian process &#x0028 ESGP &#x0029 to design a tracking controller for a PMA-driven lower limb exoskeleton. The dynamics of the system include the PMA actuation and mechanism of the leg orthoses; thus, the system is represented by two nonlinear uncertain subsystems. To facilitate the design of the controller, joint angles of leg orthoses are forecasted based on the universal approximation ability of the ESGP. A gradient descent algorithm is employed to solve the optimization problem and generate the control signal. The stability of the closed-loop system is guaranteed when the ESGP is capable of approximating system dynamics. Simulations and experiments are conducted to verify the approximation ability of the ESGP and achieve gait pattern training with four healthy subjects.