Modeling of physical activity behavioral interventions relying on MPC strategy

Salazar, Carlos, Cesar, Martin, Aguirre, Adriana, Eslambolchilar, Parisa ORCID: https://orcid.org/0000-0003-4610-1643 and Asanza, Victor 2023. Modeling of physical activity behavioral interventions relying on MPC strategy. Applied Sciences 13 (11) , 6437. 10.3390/app13116437

PDF - Published Version Available under License Creative Commons Attribution. Download (1MB)

Abstract

Physical inactivity is becoming an important threat to public health in today’s society. The COVID-19 pandemic has also reduced physical activity (PA) levels given all the restrictions imposed worldwide. In this work, physical activity interventions supported by mobile devices and relying on control engineering principles were proposed. The model was constructed relying on previous studies that consider a fluid analogy of Social Cognitive Theory (SCT), which is a psychological theory that describes how people acquire and maintain certain behaviors, including health-promoting behaviors, through the interplay of personal, environmental, and behavioral factors. The obtained model was validated using secondary data (collected earlier) from a real intervention with a group of male subjects in Great Britain. The present model was extended with new technology for a better understanding of behavior change interventions. This involved the use of applications, such as phone-based ecological momentary assessments, to collect behavioral data and the inclusion of simulations with logical reward conditions for reaching the behavioral threshold. A goal of 10,000 steps per day is recommended due to the significant link observed between higher daily step counts and lower mortality risk. The intervention was designed using a Model Predictive Control (MPC) algorithm configured to obtain a desired performance. The system was tested and validated using simulation scenarios that resemble different situations that may occur in a real setting.

Item Type:	Article
Date Type:	Publication
Status:	Published
Schools:	Computer Science & Informatics
Publisher:	Balcan Society of Geometers
ISSN:	1454-5101
Date of First Compliant Deposit:	31 May 2023
Date of Acceptance:	17 May 2023
Last Modified:	01 Jun 2023 20:24
URI:	https://orca.cardiff.ac.uk/id/eprint/160062

Actions (repository staff only)

Edit Item