Doctor of Philosophy (PhD)
First Advisor's Name
First Advisor's Committee Title
Second Advisor's Name
Kang K. Yen
Second Advisor's Committee Title
Third Advisor's Name
Third Advisor's Committee Title
Fourth Advisor's Name
Fourth Advisor's Committee Title
Fifth Advisor's Name
Fifth Advisor's Committee Title
biomedical, controls and control theory, signal processing
Date of Defense
Recent developments in prosthetics have enabled the development of powered prosthetic ankles (PPA). The advent of such technologies drastically improved impaired gait by increasing balance and reducing metabolic energy consumption by providing net positive power. However, control challenges limit performance and feasibility of today’s devices. With addition of sensors and motors, PPA systems should continuously make control decisions and adapt the system by manipulating control parameters of the prostheses. There are multiple challenges in optimization and control of PPAs. A prominent challenge is the objective setup of the system and calibration parameters to fit each subject. Another is whether it is possible to detect changes in intention and terrain before prosthetic use and how the system should react and adapt to it.
In the first part of this study, a model for energy expenditure was proposed using electromyogram (EMG) signals from the residual lower-limbs PPA users. The proposed model was optimized to minimize energy expenditure. Optimization was performed using a modified Nelder-Mead approach with a Latin Hypercube sampling. Results of the proposed method were compared to expert values and it was shown to be a feasible alternative for tuning in a shorter time.
In the second part of the study, the control challenges regarding lack of adaptivity for PPAs was investigated. The current PPA system used is enhanced with impedance-controlled parameters that allow the system to provide different assistance. However, current systems are set to a fixed value and fail to acknowledge various terrain and intentions throughout the day. In this study, a pseudo-real-time adaptive control system was proposed to predict the changes in the gait and provide a smoother gait. The proposed control system used physiological, kinetic, and kinematic data and fused them to predict the change. The prediction was done using machine learning-based methods. Results of the study showed an accuracy of up to 89.7 percent for prediction of change for four different cases.
Previously Published In
Atri, R., Marquez, J., Leung, C., Siddiquee, M., Murphy, D., Gorgey, A., Lovegreen, W., Fei, D.Y. and Bai, O., 2018. Smart DataDriven Optimization of Powered Prosthetic Ankles Using Surface Electromyography. Sensors, 18(8), p.2705.
Atri, R., Marquez, J.S., Murphy, D., Gorgey, A., Fei, D., Fox, J., Burkhardt, B., Lovegreen, W. and Bai, O., 2016, December. Investigation of muscle activity during loaded human gait using signal processing of multi-channel surface EMG and IMU. In Signal Processing in Medicine and Biology Symposium (SPMB), 2016 IEEE (pp. 1-6). IEEE.
Bai, O., Atri, R., Marquez, J.S. and Fei, D.Y., 2017, March. Characterization of lower limb activity during gait using wearable, multichannel surface EMG and IMU sensors. In Electrical Engineering Congress (iEECON), 2017 International (pp. 1-4). IEEE.
Atri, R., Marquez, J.S., Murphy, D., Gorgey, A., Fei, D., Fox, J., Lovegreen, W. and Bai, O., 2017, July. EMG-based energy expenditure optimization for active prosthetic leg tuning. In Engineering in Medicine and Biology Society (EMBC), 2017 39th Annual International Conference of the IEEE (pp. 394-397). IEEE.
Atri, Roozbeh, "Sensor-Based Adaptive Control and Optimization of Lower-Limb Prosthesis." (2019). FIU Electronic Theses and Dissertations. 4328.
In Copyright. URI: http://rightsstatements.org/vocab/InC/1.0/
This Item is protected by copyright and/or related rights. You are free to use this Item in any way that is permitted by the copyright and related rights legislation that applies to your use. For other uses you need to obtain permission from the rights-holder(s).