Closed-Loop Neural Network-Based NMES Control for Human Limb Tracking

Sharma, Nitin; Gregory, Chris; Johnson, Marcus; Dixon, Warren E.

doi:10.1109/tcst.2011.2125792

Cited by 86 publications

(48 citation statements)

References 50 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recently, efforts have been made to develop nonlinear control techniques [5], [15]- [19] to obtain improved error performance and/or asymptotic tracking [5], [17], [18] (proven through a closed-loop stability analysis), even in the presence of bounded exogenous disturbances [17], [18]. However, several open challenges persist due to the uncertain nonlinear musculoskeletal dynamics, time-dependent muscle force reduction due to fatigue, unmodeled disturbances such as muscle spasticity or external changes in muscle loads, delayed muscle response, etc.…”

mentioning

confidence: 99%

Predictor-Based Compensation for Electromechanical Delay During Neuromuscular Electrical Stimulation

Sharma

Gregory

Dixon

2011

IEEE Trans. Neural Syst. Rehabil. Eng.

View full text Add to dashboard Cite

Abstract-Electromechanical delay (EMD) is a biological artifact that arises due to a time lag between electrical excitation and tension development in a muscle. EMD is known to cause degraded performance and instability during neuromuscular electrical stimulation (NMES). Compensating for such input delay is complicated by the unknown nonlinear muscle force-length and muscle force-velocity relationships. This paper provides control development and a mathematical stability analysis of a NMES controller with a predictive term that actively accounts for EMD. The results are obtained through the development of a novel predictor-type method to address the delay in the voltage input to the muscle. Lyapunov-Krasovskii functionals are used within a Lyapunov-based stability analysis to prove semi-global uniformly ultimately bounded tracking. Experiments on able-bodied volunteers illustrate the performance and robustness of the developed controller during a leg extension trajectory following task.

show abstract

mentioning

confidence: 99%

Predictor-Based Compensation for Electromechanical Delay During Neuromuscular Electrical Stimulation

Sharma

Gregory

Dixon

2011

IEEE Trans. Neural Syst. Rehabil. Eng.

View full text Add to dashboard Cite

show abstract

“…NMES artificially contracts muscle groups by applying external stimulation current. Closed-loop control of NMES can be used to generate advanced functional tasks (in this case NMES is also called functional electrical stimulation (FES)) such as walking [2,3], single leg extension [4][5][6][7][8][9], and upper extremity grasping and reaching [10][11][12].…”

Section: List Of Tablesmentioning

confidence: 99%

“…Examples of linear control methods include proportional integral derivative (PD/PID), linear quadratic Gaussian control, pole placement method, gain scheduling control method [13][14][15]. Nonlinear control methods [4][5][6][7][8][9][16][17][18] have also been recently developed for NMES. These nonlinear methods have improved performance over linear control methods.…”

Section: List Of Tablesmentioning

confidence: 99%

“…These nonlinear methods have improved performance over linear control methods. Robust nonlinear control of NMES is especially more relevant, and controllers have been developed to compensate for the nonlinear and uncertain muscles dynamics [5,7], time-varying phenomena such as muscle fatigue [19], and electromechanical delay [18,19].…”

Section: List Of Tablesmentioning

confidence: 99%

“…These nonlinear methods were proved to have enhanced performance over linear control methods in previous researches [4][5][6][7][8][9][16][17][18]. Among these nonlinear control techniques, robust nonlinear control of NMES in [5,7] is especially more relevant given the nonlinear and uncertain muscle dynamics, time-varying phenomena such as muscle fatigue [19], and electromechanical delay [18,19].…”

Section: Figure 1 Ergys 3 Rehabilitation Systemmentioning

confidence: 99%

See 2 more Smart Citations

Robust compensation of electromechanical delay during neuromuscular electrical stimulation of antagonistic muscles

Qiu

Alibeji

Sharma

2016

2016 American Control Conference (ACC)

View full text Add to dashboard Cite

Predictor-based tracking for neuromuscular electrical stimulation

Karafyllis

Malisoff

Queiroz

et al. 2014

Int. J. Robust. Nonlinear Control

View full text Add to dashboard Cite

We present a new tracking controller for neuromuscular electrical stimulation (NMES), which is an emerging technology that artificially stimulates skeletal muscles to help restore functionality to human limbs. The novelty of our work is that we prove that the tracking error globally asymptotically and locally exponentially converges to zero for any positive input delay, coupled with our ability to satisfy a state constraint imposed by the physical system. Also, our controller only requires sampled measurements of the states instead of continuous measurements and allows perturbed sampling schedules, which can be important for practical purposes. Our work is based on a new method for constructing predictor maps for a large class of timevarying systems, which is of independent interest. NOTATION AND DEFINITIONSFor each vector x 2 R n , we let jxj denote its usual Euclidean norm, and x 0 is its transpose. The norm jMj of a matrix M 2 R m n is defined by jMj D max ¹jMxj W x 2 R n ; jxj D 1º. We let Z C denote the set of all non-negative integers. A partition D ¹T i º 1 i D0 of OE0; C1/ is any increasing sequence of times such that T 0 D 0 and T i ! C1. For every real x > 0, we let OEx denote its integer part, that is, OEx D max ¹k 2 Z C W k 6 xº. An increasing continuous function W OE0; C1/ ! OE0; C1/ is ofBy KL, we denote the set of all continuous functions W OE0; C1/ OE0; C1/ ! OE0; C1/ such that (i) for each t > 0, the mapping . ; t/ is of class K and (ii) for each s > 0, the mapping .s; / is non-increasing and satisfies lim t !C1 .s; t/ D 0. Let m and n be any positive integers. Given any open subset A Â R n and any integer j > 0, we let C j .A/ denote the class of all functions having domain A that have continuous derivatives of order j . When we wish to restrict to functions taking values in a subset Â R m , we denote the preceding set by C j .AI /. Given x W OEa r; b/ ! R n with b > a > 0 and r > 0, we let M T r .t /x denote the 'open history' of x from t r to t , that is, M T r .t /x Á .Â/ D x.t C Â/ for all Â 2 OE r; 0/ and t 2 OEa; b/. Given any interval I Â OE0; C1/, we use L 1 .I I U / to denote the space of all measurable essentially bounded functions defined on I and taking values in U Â R m . We also set kxk r D sup Â2OE r;0/ jx.Â/j for each function x. Notice that sup Â2OE r;0 jx.Â/j is not the essential supremum but the actual supremum and that is why the quantities sup Â2OE r;0 jx.Â/j and sup Â2OE r;0/ jx.Â/j do not coincide in general. We use juj OEa;b/ to denote the essential supremum of any function u over any interval OEa; b/ in its domain. A function h W A ! R where 0 2 A Â R n is called positive definite provided h.0/ D 0 and h.x/ > 0 for all x 2 A n ¹0º. A function h W R n ! R is called radially unbounded provided that for each constant M > 0, the set ¹x 2 R n W h.x/ 6 M º is bounded or empty. For any bounded function F defined on any subset of R, we let jFj 1 denote its supremum over its entire domain.(98) ‡ Some voltage-level controllers have been designed to compensate for the unknown term ...

show abstract

Closed-Loop Neural Network-Based NMES Control for Human Limb Tracking

Cited by 86 publications

References 50 publications

Predictor-Based Compensation for Electromechanical Delay During Neuromuscular Electrical Stimulation

Predictor-Based Compensation for Electromechanical Delay During Neuromuscular Electrical Stimulation

Robust compensation of electromechanical delay during neuromuscular electrical stimulation of antagonistic muscles

Predictor-based tracking for neuromuscular electrical stimulation

Contact Info

Product

Resources

About