Motor learning: a nonlinear, irreversible and fractal process The same task 5 subjects, instructed...
1
Motor learning: a nonlinear, irreversible Motor learning: a nonlinear, irreversible and fractal process and fractal process The same task 5 subjects, instructed to oscillate on a ski simulator “as ample and frequent as possible”. Movements recorded by a VICON motion analyzer (Biometrics) . References : - Nourrit, D., Delignières, D., Caillou, N., Deschamps, T. & Lauriot, B. (2003) On discontinuities in motor learning: A longitudinal study of complex skill acquisition on a ski-simulator. Journal of Motor Behavior, 35, 151-170. - Thom, R. (1983). Paraboles et Catastrophes. coll.: Champs, ed.: Flammarion, Paris. - Teulier, C., Nourrit, D. & Delignières, D. (2006). The evolution of oscillatory behavior during learning on a ski simulator. Research Quaterly for Exercise and Sport . 77, n°3, 464-475. - Nourrit-Lucas, D., Zélic, G.; Deschamps, T., Hilpron, M. & Delignières, D.(Submitted) “Once you have learned how to ride a bicycle, you never forget!”. Persistent coordination patterns in a complex task after 10 years delay. Journal of Motor Nourrit-Lucas, D. 1 ,Tossa, A.O. 1 , Zelic, G. 2 & Delignières,D. 2 . (1)MAPMO Laboratory, University of Orléans, France; (2) Movement to Health Laboratory , University of Montpellier, France e-mail:[email protected] 0 ) , ( ) ( x x x f x g x m The persistent coordination patterns after 10 years delay A retention test: ten 1-min trials, with a 1 min break between trials, 10 years later. On discontinuities in motor learning A bi-stable transition Fractal properties for the expert : close to 1/f noise A longitudinal study of 39 sessions of ten-1 min trials . Aim: examine the qualitative behavioral reorganizations that occurred during the acquisition of a complex motor skill. Procedure: the motion of the apparatus platform modelled as a self-sustained oscillator Results: 2 stages during practice. An initial stage characterized by a highly non-linear stiffness function and a Rayleigh damping function which provides a prolonged dwelling time when subjects approach the reversal points of the movement. These behaviours contributed also to well control the reversal points. On the other hand a final stage characterized by the linearization of stiffness and a van der Pol damping function where the decrease of non-linear stiffness are revealing a better storing energy in the rubber belts and a van der Pol damping allows subjects to adopt high frequencies and large amplitude that characterize skilled performance (expert behaviour in ski- simulator). Position Stiffn ess Damping Van der Pol Rayleigh The transition between Rayleigh and van der Pol damping -1.5 0 1.5 0 50 100 150 200 250 300 350 400 Sujet 4 essais Initial stage « Rayleigh » Final stage « Van der Pol » Transition Phase Van der Pol Rayleigh Discovery stage A saddle-node bifurcation: Mono-stable phase (initial stage) Bi-stable phase (transition phase) Mono-stable phase (final stage) Practice (time) Drift before transition The non-linearity of the motor learning seems to have the form of a prolonged phase transition (like in a saddle-node bifurcation) ,which permits access to the skilled behavior. C01(Rayleigh) is the linear damping coefficient obtained with this forcing procedure. C01(Rayleigh) is negative when the limit cycle is sustained by a Rayleigh behavior, and positive for a van der Pol behavior. The exploitation of the previous damping behavior, the van der Pol damping, appeared quite durable during the time. Any transformation in the attractors landscape is “catastrophic” in sense of René Thom (1983), the space of phases is modified in an irreversible way. One allows us to validate the old saying : “Once one learns to ride a bicycle, one never forgets that skill”. Novice Exper t A comparative test between 5 experts and 5 novices during one-10 min trial. Time series of ski-simulator movements analysed by DFA. 1/f noise is an adaptation- stability trade-off : a localized perturbation will not necessarily alter the stability of the global. The expert’s time series have revealed a noise close to the pink noise. We consider the 1/f noise as an indicator of efficiency of the system.
-
Upload
diane-richardson -
Category
Documents
-
view
214 -
download
0
Transcript of Motor learning: a nonlinear, irreversible and fractal process The same task 5 subjects, instructed...
Motor learning: a nonlinear, irreversible and fractal processMotor learning: a nonlinear, irreversible and fractal process
The same task5 subjects, instructed to
oscillate on a ski simulator “as ample and frequent as
possible”. Movements recorded by a VICON
motion analyzer (Biometrics) .
References : - Nourrit, D., Delignières, D., Caillou, N., Deschamps, T. & Lauriot, B. (2003) On discontinuities in motor learning: A longitudinal study of complex skill acquisition on a ski-simulator. Journal of Motor Behavior, 35, 151-170.- Thom, R. (1983). Paraboles et Catastrophes. coll.: Champs, ed.: Flammarion, Paris.- Teulier, C., Nourrit, D. & Delignières, D. (2006). The evolution of oscillatory behavior during learning on a ski simulator. Research Quaterly for Exercise and Sport. 77, n°3, 464-475. - Nourrit-Lucas, D., Zélic, G.; Deschamps, T., Hilpron, M. & Delignières, D.(Submitted) “Once you have learned how to ride a bicycle, you never forget!”. Persistent coordination patterns in a complex task after 10 years delay. Journal of Motor Behavior.
.
Nourrit-Lucas, D.1,Tossa, A.O.1, Zelic, G. 2 & Delignières,D. 2. (1)MAPMO Laboratory, University of Orléans, France; (2) Movement to Health Laboratory , University of Montpellier, France
e-mail:[email protected]
0),()( xxxfxgxm
The persistent coordination patterns after 10 years delayA retention test: ten 1-min trials, with a 1 min break between trials, 10 years later.
On discontinuities in motor learning A bi-stable transition
Fractal properties for the expert : close to 1/f noise
A longitudinal study of 39 sessions of ten-1 min trials . Aim: examine the qualitative behavioral reorganizations that occurred during the acquisition of a complex motor skill. Procedure: the motion of the apparatus platform modelled as a self-sustained oscillator
Results: 2 stages during practice. An initial stage characterized by a highly non-linear stiffness function and a Rayleigh damping function which provides a prolonged dwelling time when subjects approach the reversal points of the movement. These behaviours contributed also to well control the reversal points. On the other hand a final stage characterized by the linearization of stiffness and a van der Pol damping function where the decrease of non-linear stiffness are revealing a better storing energy in the rubber belts and a van der Pol damping allows subjects to adopt high frequencies and large amplitude that characterize skilled performance (expert behaviour in ski-simulator).
Position Stiffness Damping
Van der Pol
Rayleigh
The transition between Rayleigh and van der Pol damping
-1.5
0
1.5
0 50 100 150 200 250 300 350 400
Sujet 4
essais
Initial stage « Rayleigh »
Final stage « Van der Pol »
Transition Phase
Van der Pol
Rayleigh
Discovery stage
A saddle-node bifurcation:
Mono-stable phase(initial stage)
Bi-stable phase (transition phase)
Mono-stable phase (final stage)
Practice (time)
Drift before transitionThe non-linearity of the motor learning seems to have the form of a prolonged phase transition (like in a saddle-node bifurcation) ,which permits access to the skilled behavior.
C01(Rayleigh) is the linear damping coefficient obtained with this forcing procedure. C01(Rayleigh) is negative when the limit cycle is sustained by a Rayleigh behavior, and positive for a van der Pol behavior.
The exploitation of the previous damping behavior, the van der Pol damping, appeared quite durable during the time. Any transformation in the attractors landscape is “catastrophic” in sense of René Thom (1983), the space of phases is modified in an irreversible way. One allows us to validate the old saying : “Once one learns to ride a bicycle, one never forgets that skill”.
Novice Expert
A comparative test between 5 experts and 5 novices during one-10 min trial.
Time series of ski-simulator movements analysed by DFA.
1/f noise is an adaptation- stability trade-off : a localized perturbation will not necessarily alter the stability of the global. The expert’s time series have revealed a noise close to the pink noise. We consider the 1/f noise as an indicator of efficiency of the system.
