Musculoskeletal Neural Network path generator for a virtual upper-limb active controlled orthosis

In this paper, a non-parametric model of the neuromusculoskeletal system for the biceps brachii is presented. The model serves to generate angular paths for the control of a virtual active orthosis. The path generator uses a differential neural network (DNN) identifier that obtains the reference angular position and velocities using the raw electromyographic (EMG) signals as input. The model is validated using experimental data. The training and closed-loop implementation of the proposed model are described. The control strategy ensures that the user reaches a set-point with a predefined position constraint and that the device follows the natural reference path that corresponds to the raw EMG signal.

[1]  Suncheol Kwon,et al.  Real-Time Upper Limb Motion Estimation From Surface Electromyography and Joint Angular Velocities Using an Artificial Neural Network for Human–Machine Cooperation , 2011, IEEE Transactions on Information Technology in Biomedicine.

[2]  Neil E. Cotter,et al.  The Stone-Weierstrass theorem and its application to neural networks , 1990, IEEE Trans. Neural Networks.

[3]  Andrew J. Pullan,et al.  Neuromuscular Interfacing: Establishing an EMG-Driven Model for the Human Elbow Joint , 2012, IEEE Transactions on Biomedical Engineering.

[4]  David G Lloyd,et al.  Neuromusculoskeletal modeling: estimation of muscle forces and joint moments and movements from measurements of neural command. , 2004, Journal of applied biomechanics.

[5]  J. Soucie,et al.  Range of motion measurements: reference values and a database for comparison studies , 2011, Haemophilia : the official journal of the World Federation of Hemophilia.

[6]  Alexander S. Poznyak,et al.  Differential Neural Networks for Robust Nonlinear Control , 2004, IEEE Transactions on Neural Networks.

[7]  Avrie Levent,et al.  Robust exact differentiation via sliding mode technique , 1998, Autom..

[8]  Alexander S. Poznyak,et al.  Adaptive sliding-mode control with integral compensation for robotic devices with state constraints , 2018 .

[9]  Leonid M. Fridman,et al.  Second-order sliding-mode observer for mechanical systems , 2005, IEEE Transactions on Automatic Control.

[10]  Vadim I. Utkin,et al.  Decentralized sliding-mode control of robotic manipulator with constraint workspace: a finite-convergent barrier Lyapunov approach , 2019, 2019 16th International Conference on Electrical Engineering, Computing Science and Automatic Control (CCE).

[11]  Namig J. Guliyev,et al.  On the Approximation by Single Hidden Layer Feed-forward Neural Networks With Fixed Weights , 2017, Neural Networks.

[12]  John Whyte,et al.  The Importance of Voluntary Behavior in Rehabilitation Treatment and Outcomes. , 2019, Archives of physical medicine and rehabilitation.

[13]  Yasuharu Koike,et al.  Estimation of dynamic joint torques and trajectory formation from surface electromyography signals using a neural network model , 1995, Biological Cybernetics.

[14]  Isaac Chairez,et al.  Integrated wearable and self-carrying active upper limb orthosis , 2018, Proceedings of the Institution of Mechanical Engineers. Part H, Journal of engineering in medicine.

[15]  Isaac Chairez,et al.  Super-twisting sliding mode differentiation for improving PD controllers performance of second order systems. , 2014, ISA transactions.

[16]  George Cybenko,et al.  Approximation by superpositions of a sigmoidal function , 1989, Math. Control. Signals Syst..

[17]  Robert F. Stengel,et al.  Smooth function approximation using neural networks , 2005, IEEE Transactions on Neural Networks.