PreprintPDF Available

A New Weight Compensation Model Considering Joint Misalignments Monitored by a Feedforward Impedance Control Method Applied To an Active Upper-Limb Exoskeleton

  • August 2021
DOI:10.21203/rs.3.rs-786361/v1
Authors:
Dorian Verdel at Imperial College London
Dorian Verdel
Simon Bastide at Université Paris-Saclay
Simon Bastide
Nicolas Vignais at Université de Rennes 2
Nicolas Vignais
Olivier Bruneau at Université Paris-Saclay
Olivier Bruneau
Show all 5 authorsHide
Download file PDF
Read file
Download citation

Abstract and Figures

Background Active exoskeletons are promising devices for improving rehabilitation procedures in patients. In particular, exoskeletons implementing human limb's weight support (WS) are interesting to restore some mobility in patients with muscle weakness. Using active exoskeletons should result in accurate and generic WS but its effect on human motor control will critically depend on the position of the user within the exoskeleton and the characteristics of the control law. Methods The present study aims at improving WS of the upper limb by providing a weight model considering joint misalignments and a control law including feedforward terms learned from a prior population-based analysis. Three experiments are respectively conducted on 29, 17 and 19 participants who performed posture maintenance and pointing movements with the forearm in the sagittal plane. The first two experiments were used to build an accurate WS control law and the third experiment was conducted to compare the effects of different WS control laws on human movement and assess their quality. During these three experiments, kinematic data and eletromyographic activity of elbow flexors and extensors were measured. Interaction forces were measured with a force/torque sensor placed between the human segment and the robot link. Results The introduction of joint misalignments in the WS model allowed to drastically reduce the model errors in terms of weight estimation. The use of a feedforward architecture based on model and errors learned during experiments, coupled to a force feedback, allowed to reduce model tracking errors in both static and dynamic conditions during vertical movements, which induce substantial variations of gravitational torques. Overall, WS did not affect the general kinematic motion parameters of the participants and decreased the activity of antigravity muscles (flexors). However, WS increased the activation of extensors because weight is usually exploited by humans to accelerate a limb downward. Conclusion A new weight compensation model considering joint misalignments was introduced and data showed their prominent role on WS accuracy and homogeneity. Three WS control laws were compared and results indicated that classical control methods were not sufficient to provide an accurate tracking of the weight model during dynamic vertical movements but that introducing simple feedforward terms learned from previous measures could significantly improve WS accuracy. Accordingly, WS reduced significantly activity in flexors in both static and dynamic conditions. Nevertheless, WS tended to increase the activity of extensors, which might be an important factor in a rehabilitation perspective. Indeed, if the present WS control law will be very helpful to allow patients accelerating the arm upward despite some muscle weakness, it may have an opposite effect when accelerating the arm downward. A partial WS controller could thus be more appropriate in rehabilitation applications.
Figures - available via license: Creative Commons Attribution 4.0 International
Content may be subject to copyright.
Available via license: CC BY 4.0
Content may be subject to copyright.
y0
z0=xr×y0
kxr×y0k
x0=y0×z0
x0z0
xr
O=OrOr
(O, x0,y0)
xr=Rr0(z0, φr)x0
Rr0(z0, φr)φr
z0
Rab(v, θ)Rab
b a v
θ
yr=Rr0(z0, φr)y0
zr=Rr0(z0, φr)z0=z0
xh
yh=yh,pxh
xh,p
xh
yh,p=Rhr,1(z0, φh(φr)) yr
xh,p=Rhr,1(z0, φh(φr)) xr
yh=yh,p
xh=Rhr,2(yh,p, ψh(φr)) xh,p
xhxr
φr
xh(O, x0,y0)
φh(φr)
Rhr,2(yh,pψh(φr))
Wh
h=Wh,p
h=Rh0(z0, φr+φh(φr))TW0
h
Wh
hW0
h
φrφh(φr)
zr
(Oh,p,xh,p)
xh
(O, x0,y0)
xr
yr
φr
φr=q
q
τhq
τh= (xgxh×Wh)·z0
=−mf,h gxgcos(q+φh(q))
τh
(Oh,xh)mf,h
xg
q
φh(q)
F=mf,h
xg
li
gcos(q+φh(q))
cos(φh(q))
F
li
xh
(Oh,xh)
Or
Or
τi=lF
=lmf,h
xg
li
gcos(q+φh(q))
cos(φh(q))
τi
l
Or
φhφh,0
τi,wjm =lmf,h
xg
li
gcos(q)
τi,wjm
φh(q)
φh(q)≈φh
φh+q > 40◦φh+q < −60◦
φh(q)
ˆmeq,h
ˆ
φh= arg min
meq,h,φhF−meq ,hgcos(q+φh)
cos(φh)2
ˆmeq,h ˆ
φhmeq,h
φhmeq,h =mf,h xg/li
F q
17
7 10
70.7±9.7kg 176 ±7.8cm
24 ±4.9
25.5±1.4cm 18.6±1.3cm
2017 34
20
[−85.94◦,39.53◦]
5s
1kHz
φ,0
MAE =1
NX|ˆ
F(q, φ)−F|
Nˆ
F(q, φ)
F
t
p < 0.05
τi=αlmeq,h gcos(q+φh)
cos(φh)
α
α
1 0
τctrl =Kp(ˆτi(q, φ)−τi)
+KiZt
0
(ˆτi(q, φ)−τi)dt + ˆτcomp (¨q, ˙q, q)
τctrl
ˆτi(q, φ)
τi
ˆτcomp (¨q, ˙q, q)
¨q˙q q
ˆτcomp (¨q, ˙q, q) = ˆ
M(q)¨q+ˆ
C(q, ˙q)+ ˆ
G(q)+sign( ˙q)ˆτC+ˆν˙q
ˆ
M(q)
ˆ
C(q, ˙q)
ˆ
G(q)
ˆτC
ˆν
ˆτcomp
τF F,τ =KF F,τ ˆτi(q, φ) + τF F ,τ 0
τF F,τ
KF F,τ
ˆτi(q, φ)
τF F,τ 0
l
τF F,q =KFF,q q+τF F,q 0
τF F,q
KF F,q
q
τF F,q 0
τctrl =Kp(ˆτi(q, φ)−τi)
+KiZt
0
(ˆτi(q , φ)−τi)dt + ˆτcomp ( ˙q, q)
+τF F,τ +τFF,q
α
60◦
F
N= 8
kg cm cm cm
23.2±3.0 67.1±11.8 174.7±7.6 25.5±1.6 18.8±1.0
24 ±4.9 70.7±9.7 176 ±7.8 25.5±1.4 18.6±1.3
24.3±2.2 70.7±9.8 176 ±8.4 26.4±2.0 19.5±1.2
N
meq,h (kg)φh(rad)
2.13 ±0.56 0.27 ±0.12
1.93 ±0.26 0.21 ±0.07
1.90 ±0.36 0.23 ±0.07
1010
7kHz
Tup Tbot Tmid
60◦
1s
600 ms
α= 1
2s
7 30 15
15 210
6
6
4 30
15 15
120
Tup
Tbot Tmid
3
3
τi
τi=lF
l
F
l
F τi
α= 1
τi>0Nm
t−
p < 0.05
N N
φ,0 1.9 0.68
φ  <0.62 0.28
meq,h ≈2.45 kg ;φ≈0.217 rad
meq,h ≈2.14 kg
t
T−7.18
p= 3.91 ×10−7
3
2.4
meq,h φh
meq,h kg φhrad
1.93 0.21
0.26 0.07
φh
Tup
Tbot Tmid
Tmid Tup
F2,36 = 36.7p= 0.3×10−5η2= 0.67
p= 1.1×10−4p= 0.7×10−5
p= 2.5×10−3
F
τi
27%
24%
44%
p= 2.49 ×10−11, F =
54.5t−
p= 9.3×
10−5, T = 5.61 p= 5.54 ×
10−8, T = 9.88 p=
2.85 ×10−4, T = 5.07
F3,54 = 0.446 p= 0.721
η2= 0.0242
rad/s rad/s2s
29.1±6.8 157.4±56.2 0.64 ±0.14
28.8±6.5 155.8±52.3 0.64 ±0.17
30.9±6.9 169.7±60.8 0.61 ±0.15
29.8±6.4 165.3±56.5 0.61 ±0.12
F3,54 = 553.14 p= 2.82×10−28 η2= 0.97
p= 2.2×10−5p= 2.1×10−5p= 9.5×10−5
F3,54 = 3.73 p= 0.046 η2= 0.17
F3= 39.82
p= 3.52 ×10−8η2= 0.69
p=
1.01 ×10−7p= 4.49 ×10−5p= 4.99 ×10−8
F3= 41.52
p= 4.73 ×10−14 η2= 0.70
p= 5.44 ×10−7p= 3.25 ×10−9p= 1.46 ×10−5
Rr0(z0, φr) =
cos(φr)−sin(φr) 0
sin(φr) cos(φr) 0
0 0 1
Rhr,1(z0, φh(φr)) =
cos(φh(φr)) −sin(φh(φr)) 0
sin(φh(φr)) cos(φh(φr)) 0
0 0 1
Rhr,2(yh,p, ψh(φr)) =
cos(ψh(φr)) 0 sin(ψh(φr))
0 1 0
−sin(ψh(φr)) 0 cos(ψh(φr))
Rh0(z0, φr+φh(φr)) =
cos(φr+φh(φr)) −sin(φr+φh(φr)) 0
sin(φr+φh(φr)) cos(φr+φh(φr)) 0
0 0 1
ResearchGate has not been able to resolve any citations for this publication.
A cross-species neural integration of gravity for motor optimization
Article
Full-text available
  • Apr 2021
Recent kinematic results, combined with model simulations, have provided support for the hypothesis that the human brain shapes motor patterns that use gravity effects to minimize muscle effort. Because many different muscular activation patterns can give rise to the same trajectory, here, we specifically investigate gravity-related movement properties by analyzing muscular activation patterns during single-degree-of-freedom arm movements in various directions. Using a well-known decomposition method of tonic and phasic electromyographic activities, we demonstrate that phasic electromyograms (EMGs) present systematic negative phases. This negativity reveals the optimal motor plan’s neural signature, where the motor system harvests the mechanical effects of gravity to accelerate downward and decelerate upward movements, thereby saving muscle effort. We compare experimental findings in humans to monkeys, generalizing the Effort-optimization strategy across species.
View
An Identification-Based Method Improving the Transparency of a Robotic Upper Limb Exoskeleton
Article
Full-text available
  • Feb 2021
Over the past decade, research on human–robot collaboration has grown exponentially, motivated by appealing applications to improve the daily life of patients/operators. A primary requirement in many applications is to implement highly “transparent” control laws to reduce the robot impact on human movement. This impact may be quantified through relevant motor control indices. In this paper, we show that control laws based on careful identification procedures improve transparency compared to classical closed-loop position control laws. A new performance index based on the ratio between electromyographic activity and limb acceleration is also introduced to assess the quality of human exoskeleton interaction.
View
Human arm weight compensation in rehabilitation robotics: efficacy of three distinct methods
Article
Full-text available
Background: Arm weight compensation with rehabilitation robots for stroke patients has been successfully used to increase the active range of motion and reduce the effects of pathological muscle synergies. However, the differences in structure, performance, and control algorithms among the existing robotic platforms make it hard to effectively assess and compare human arm weight relief. In this paper, we introduce criteria for ideal arm weight compensation, and furthermore, we propose and analyze three distinct arm weight compensation methods (Average, Full, Equilibrium) in the arm rehabilitation exoskeleton ’ARMin’. The effect of the best performing method was validated in chronic stroke subjects to increase the active range of motion in three dimensional space. Conclusion: Different arm weight compensation methods were developed according to initially defined criteria. The methods were then analyzed with respect to their sensitivity and required technology. In general, weight compensation performance improved with the level of technology, but increased cost and calibration efforts. This study reports a systematic way to analyze the efficacy of different weight compensation methods using EMG. Additionally, the feasibility of the best method, Equilibrium, was shown by testing with three stroke patients. In this test, a height dependence of the workspace size also seemed to be present, which further highlights the importance of patient-specific weight compensation, particularly for training at different arm heights.
View
Movement kinematics and proprioception in post-stroke spasticity: assessment using the Kinarm robotic exoskeleton
Article
Full-text available
Background: Motor impairment after stroke interferes with performance of everyday activities. Upper limb spasticity may further disrupt the movement patterns that enable optimal function; however, the specific features of these altered movement patterns, which differentiate individuals with and without spasticity, have not been fully identified. This study aimed to characterize the kinematic and proprioceptive deficits of individuals with upper limb spasticity after stroke using the Kinarm robotic exoskeleton. Methods: Upper limb function was characterized using two tasks: Visually Guided Reaching, in which participants moved the limb from a central target to 1 of 4 or 1 of 8 outer targets when cued (measuring reaching function) and Arm Position Matching, in which participants moved the less-affected arm to mirror match the position of the affected arm (measuring proprioception), which was passively moved to 1 of 4 or 1 of 9 different positions. Comparisons were made between individuals with (n = 35) and without (n = 35) upper limb post-stroke spasticity. Results: Statistically significant differences in affected limb performance between groups were observed in reaching-specific measures characterizing movement time and movement speed, as well as an overall metric for the Visually Guided Reaching task. While both groups demonstrated deficits in proprioception compared to normative values, no differences were observed between groups. Modified Ashworth Scale score was significantly correlated with these same measures. Conclusions: The findings indicate that individuals with spasticity experience greater deficits in temporal features of movement while reaching, but not in proprioception in comparison to individuals with post-stroke motor impairment without spasticity. Temporal features of movement can be potential targets for rehabilitation in individuals with upper limb spasticity after stroke.
View
Effects of arm weight support on neuromuscular activation during reaching in chronic stroke patients
Article
Full-text available
To better understand how arm weight support (WS) can be used to alleviate upper limb impairment after stroke, we investigated the effects of WS on muscle activity, muscle synergy expression, and corticomotor excitability (CME) in 13 chronic stroke patients and 6 age-similar healthy controls. For patients, lesion location and corticospinal tract integrity were assessed using magnetic resonance imaging. Upper limb impairment was assessed using the Fugl-Meyer upper extremity assessment with patients categorised as either mild or moderate–severe. Three levels of WS were examined: low = 0, medium = 50 and high = 100% of full support. Surface EMG was recorded from 8 upper limb muscles, and muscle synergies were decomposed using non-negative matrix factorisation from data obtained during reaching movements to an array of 14 targets using the paretic or dominant arm. Interactions between impairment level and WS were found for the number of targets hit, and EMG measures. Overall, greater WS resulted in lower EMG levels, although the degree of modulation between WS levels was less for patients with moderate–severe compared to mild impairment. Healthy controls expressed more synergies than patients with moderate–severe impairment. Healthy controls and patients with mild impairment showed more synergies with high compared to low weight support. Transcranial magnetic stimulation was used to elicit motor-evoked potentials (MEPs) to which stimulus–response curves were fitted as a measure of corticomotor excitability (CME). The effect of WS on CME varied between muscles and across impairment level. These preliminary findings demonstrate that WS has direct and indirect effects on muscle activity, synergies, and CME and warrants further study in order to reduce upper limb impairment after stroke.
View
Interacting with a “Transparent” Upper-Limb Exoskeleton: A Human Motor Control Approach
Conference Paper
Full-text available
  • Oct 2018
Establishing a symbiotic relationship between a human and a exoskeleton is the end goal in many applications in order to provide benefits to the user. However, the literature focusing on the human side of human-exoskeleton interaction has remained less exhaustive than the literature focusing on the design (hardware/software) of the exoskeleton device itself. It is, though, essential to understand how a human adapts his motor control when interacting with an exoskeleton. Motor adaptation is an implicit process carried out by the central nervous system when the body encounters a perturbation, a paradigm that has been extensively studied in the field of human motor control research. When wearing an exoskeleton, even “as-transparent as- possible”, contact/interaction forces may impact well-known motor control laws in a way that may be detrimental to the user, and even compromise usability in real applications. The present paper investigates how interaction with a backdrivable upper-limb exoskeleton (ABLE) set in “transparent” mode of control affects the kinematics/dynamics of human movement in a simple task. We find that important motor control features are preserved when moving with ABLE but an overall movement slowness occurs, likely as a response to increased inertia according to optimal control simulations. Such a human motor control approach illustrates one possible way to assess the degree of symbiosis between human and exoskeleton, i.e. by grounding on well-known findings in motor control research.
View
Pingouin: statistics in Python
Article
Full-text available
  • Nov 2018
View
Progress and Prospects of the Human-Robot Collaboration
Article
Full-text available
Recent technological advances in hardware design of the robotic platforms enabled the implementation of various control modalities for improved interactions with humans and unstructured environments. An important application area for the integration of robots with such advanced interaction capabilities is human-robot collaboration. This aspect represents high socioeconomic impacts and maintains the sense of purpose of the involved people, as the robots do not completely replace the humans from the work process. The research community's recent surge of interest in this area has been devoted to the implementation of various methodologies to achieve intuitive and seamless human-robot-environment interactions by incorporating the collab-orative partners' superior capabilities, e.g. human's cogni-tive and robot's physical power generation capacity. In fact, the main purpose of this paper is to review the state-of-the-art on intermediate human-robot interfaces (bi-directional), robot control modalities, system stability, benchmarking and relevant use cases, and to extend views on the required future developments in the realm of human-robot collaboration. Arash Ajoudani is with HRI
View
Adaptive Proxy-Based Controller of an Active Ankle Foot Orthosis to Assist Lower Limb Movements of Paretic Patients
Article
  • Mar 2019
This paper deals with the control of an active ankle foot orthosis (AAFO) for paretic patients. State of the art methods using an AAFO try to track a predefined trajectory of the ankle joint while guaranteeing the wearer’s safety in the presence of a large tracking error. Combining the wearer’s safety and tracking accuracy is generally difficult to achieve at the same time, hence a trade-off should be found. Proxy-based sliding mode control (PSMC) offers great performances in both position tracking and safety guarantee. However, its tracking performance is subject to the influences of parameter uncertainties and external disturbances that generally occur during walking. This paper introduces an adaptation interaction method to the basic PSMC with an online adaptation of the proportional, integral and derivative parameters. At the same time, a gait phase-based ankle reference generation algorithm was proposed to adjust the joint reference trajectory in real time. The experiments using the AAFO show better tracking results with respect to basic PSMC while guaranteeing the safety.
View
Physiological consequences of using an upper limb exoskeleton during manual handling tasks
Article
This study aimed to assess the physiological consequences of using an upper limb exoskeleton during manual handling task, as muscle activity, upper limb kinematics, postural balance and cardiac cost. Participants performed three tasks (load lifting (LIFT), carrying (WALK) and stacking-unstacking (STACK)) with (EXOS) and without (FREE) an exoskeleton. During LIFT and STACK, the activity of the deltoid anterior muscle was significantly lower for EXOS than for FREE. During LIFT, the activity of the triceps brachii (TB) and tibialis anterior muscles significantly increased for EXO. The TB muscle activity significantly decreased for EXOS during WALK. The cardiac cost tended to increase with the use of the exoskeleton during LIFT, compared to FREE. The upper limb kinematics significantly differed between the EXOS and FREE conditions for all tasks. The benefits of the upper limb exoskeleton to reduce shoulder flexor muscle activity has been demonstrated, while broader physiological consequences have also been evidenced as increased antagonist muscle activity, postural strains, cardiovascular demand, and modified kinematics.
View