Skip to main navigation Skip to search Skip to main content

Contrôle haptique d'un tapis roulant orthopédique

Translated title of the thesis: Haptic control system for orthopedic treadmill
  • Simon Marcoux

Student thesis: Master's thesisMaster in Engineering: Engineering

Abstract

Goals: The goal of this project is to design a predictive haptic control algorithm for a split belt treadmill to achieve treadmill-on-demand capabilities. A subject will therefore be able to dictate the speed of the treadmill simply by using the ground reaction forces (GRF). Such a system will pave the way for different research projects in rehabilitation including orthopedic (gonarthrosis) and neurologic (hemiparesis) rehabilitation. Methodology: A predictive speed determination model was developed in the Matlab Simulink environment and a validation process was performed using control values to assess the symmetric and asymmetric prediction capabilities of the model. After that, the model was implemented on a split belt AMTI orthopedic treadmill. A total of six (6) subjects participated in the haptic treadmill’s validation process. Participants were also required to answer a qualitative form to get general feedback from them. The average speeds on the treadmill were compared with the average speed obtained over ground to study if subjects were able to emulate their natural walking speed on the haptic treadmill. Results: The speed determination model was able to predict requested speed with a precision of 3.5% RMS. Furthermore, an analysis of variance (ANOVA) didn’t show any statistical difference between the model and the control method used as a comparison basis. It was also observed that healthy subjects exposed to large amount of asymmetry will significantly alter their gait pattern. The real time implementation of the model allowed subjects to control on demand the treadmill’s speed. Subjects also reported that the overall functionality of the treadmill was intuitive and safe. Conclusion: This research project designed a haptic control algorithm for a split belt orthopedic treadmill and was able to characterize the performance of the underlying mathematical model. It was also proved that subjects were able to intuitively control the treadmill. With such a tool in hand, it would be interesting to pursue further research endeavor in rehabilitation and to continue the development and characterization of the system through longer clinical tryouts.
Date20 May 2015
Original languageFrench
Awarding Institution
  • École de technologie supérieure
SupervisorRachid Aissaoui (Supervisor)

Cite this

'