Research Papers

A Neuronal Model of Central Pattern Generator to Account for Natural Motion Variation

[+] Author and Article Information
Reza Sharif Razavian

Motion Research Group,
Systems Design Engineering,
University of Waterloo,
Waterloo, ON N2L 3G1, Canada
e-mail: rsharifr@uwaterloo.ca

Naser Mehrabi

Motion Research Group,
Systems Design Engineering,
University of Waterloo,
Waterloo, ON N2L 3G1, Canada
e-mail: nmehrabi@uwaterloo.ca

John McPhee

Fellow ASME
Motion Research Group,
Systems Design Engineering,
University of Waterloo,
Waterloo, ON N2L 3G1, Canada
e-mail: mcphee@uwaterloo.ca

Contributed by the Design Engineering Division of ASME for publication in the JOURNAL OF COMPUTATIONAL AND NONLINEAR DYNAMICS. Manuscript received January 23, 2015; final manuscript received July 7, 2015; published online August 26, 2015. Assoc. Editor: Ahmet S. Yigit.

J. Comput. Nonlinear Dynam 11(2), 021007 (Aug 26, 2015) (9 pages) Paper No: CND-15-1019; doi: 10.1115/1.4031086 History: Received January 23, 2015

We have developed a simple mathematical model of the human motor control system, which can generate periodic motions in a musculoskeletal arm. Our motor control model is based on the idea of a central pattern generator (CPG), in which a small population of neurons generates periodic limb motion. The CPG model produces the motion based on a simple descending command—the desired frequency of motion. Furthermore, the CPG model is implemented by a spiking neuron model; as a result of the stochasticity in the neuron activities, the motion exhibits a certain level of variation similar to real human motion. Finally, because of the simple structure of the CPG model, it can generate the sophisticated muscle excitation commands much faster than optimization-based methods.

Copyright © 2016 by ASME
Your Session has timed out. Please sign back in to continue.



Grahic Jump Location
Fig. 1

Schematic of the musculoskeletal forearm model

Grahic Jump Location
Fig. 2

The layered structure of the CPG controller

Grahic Jump Location
Fig. 3

(a) An example for the average firing rate of 100 neurons to a changing stimulus. (b) The synaptic weight between the 100 neurons is optimally calculated, so that the weighted sum of the neurons' firing rate represents the square of the input stimulus.

Grahic Jump Location
Fig. 4

Schematic of CPG controller implementation with spiking neurons

Grahic Jump Location
Fig. 5

The reference elbow angle, θdes; three periods of motion are shown

Grahic Jump Location
Fig. 6

The optimization framework to find the Fourier series parameters, which will be used in the online generation of muscle excitation signals

Grahic Jump Location
Fig. 7

Optimal Fourier series coefficients for BRD muscle at different periods of motion

Grahic Jump Location
Fig. 8

Comparison of the experimental data with the CPG model response. Left column compares the muscle excitation patterns in one cycle with the average experimental EMGs. Right column compares the resulting motion between the model and experiments. The results are shown for two speeds of motion: (a) fast motion, T = 1.5 s and (b) slow motion, T = 3 s.




Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In