by Ricardo Bautista-Quintero, Rickey Dubay, Juan A Carretero, Miguel Diaz-Rodriguez
Abstract:
In the last decade, technology related to powered prosthetic legs has demonstrated solid potential towards returning fully functional capabilities of millions of lower-limb-amputees. These remarkable advantages in Prosthetic devices have integrated state-of-the-art mechatronic systems which are able to receive and execute electric-signal commands of the remaining leg. However, replicating the inherent complexity of human mobility requires a robust and adaptive control strategy which must be capable to control prostheticactuators efficiently and be clinically viable. In this context, the paper introduces a control approach based on General Predictive Control (GPC) + Integral compensator. This technique is based on a prediction mechanism that optimally minimizes the sum of the square errors between the future prediction and current measurements. Although robustness of the optimal GPC have been widely evaluated for linear systems, this approach uses a nonlinear model of a Two-Degrees-of-Freedom prosthetic leg. A model identification algorithm estimates the transfemoral device dynamic behaviour in order to design a GPC control partially decoupled in two decentralized subsystems. An integral compensator is added to reduce the uncertainties due to residual dynamic caused for both links interaction and friction torques in the joints.
Reference:
Pilot study of general predictive control+ integral compensator for powered prosthetic legs (Ricardo Bautista-Quintero, Rickey Dubay, Juan A Carretero, Miguel Diaz-Rodriguez), In 2015 CCToMM Symposium on Mechanisms, Machines, and Mechatronics, 2015.
Bibtex Entry:
@inproceedings{bautista2015b,
title={Pilot study of general predictive control+ integral compensator for powered prosthetic legs},
author={Bautista-Quintero, Ricardo and Dubay, Rickey and Carretero, Juan A and Diaz-Rodriguez, Miguel},
booktitle={2015 CCToMM Symposium on Mechanisms, Machines, and Mechatronics},
year={2015},
location= {Ottawa},
eventdate = {2015-05-28/2015-05-29},
url={https://www.researchgate.net/publication/278785431_PILOT_STUDY_OF_GENERAL_PREDICTIVE_CONTROL__INTEGRAL_COMPENSATOR_FOR_POWERED_PROSTHETIC_LEGS?ev=prf_pub},
abstract={In the last decade, technology related to powered prosthetic legs has demonstrated solid potential towards returning fully functional capabilities of millions of lower-limb-amputees. These remarkable advantages in Prosthetic devices have integrated state-of-the-art mechatronic systems which are able to receive and execute electric-signal commands of the remaining leg. However, replicating the inherent complexity of human mobility requires a robust and adaptive control strategy which must be capable to control prostheticactuators efficiently and be clinically viable. In this context, the paper introduces a control approach based on General Predictive Control (GPC) + Integral compensator. This technique is based on a prediction mechanism that optimally minimizes the sum of the square errors between the future prediction and current measurements. Although robustness of the optimal GPC have been widely evaluated for linear systems, this approach uses a nonlinear model of a Two-Degrees-of-Freedom prosthetic leg. A model identification algorithm estimates the transfemoral device dynamic behaviour in order to design a GPC control partially decoupled in two decentralized subsystems. An integral compensator is added to reduce the uncertainties due to residual dynamic caused for both links interaction and friction torques in the joints.},
keywords={Power Prosthetic Leg; GPC+Integral Approach; System Identification, Non-Linear Model,
Mechatronic system},
}