Plateforme multi-capteurs pour le monitorage du positionnement geographique et des signaux comportementaux.

摘要

Nowadays, people are more and more touched by chronic diseases. Diabetes and cardiovascular diseases are of the number. Obesity, considered both as a disease and a risk factor for other illnesses, is also a public health problem. In order to study the causes and consequences of these diseases, a large number of studies are conducted to examine the patients. These studies’ goals are to get data on some physiological parameters and on the environment in which the patients live. One of the big questions is to know how and how much does the environment and the way of living affect the emergence and increase of these diseases.;In order to be able to do these studies, more and more sophisticated tools are used to measure and get the relevant data. Commercial devices can be used, increasing the number of devices the participant of the study has to wear. The new capabilities of smart phones, bearing a lot of different sensors, could help reduce the burden of the participants. However a lot of problems concerning the battery lifetime and the lack of precision of some data have been observed. This is the main reason why more and more research groups start to develop themselves custom systems in order to fulfill their needs.;This thesis present the participation to the design and development of such custom system that can get both the geographic position and physiologic data on a patient, while being portable, not cumbersome and being able to work continuously for more than 12 hours. In order to be able to work that long, optimizations have been made in the battery management. A strategy is to determine when, during the study, some data are not useful and to shut down the corresponding modules in the device, thus allowing to preserve the battery much longer.;In order to verify the impact of the different smart battery management algorithms a few tests have been conducted. Two units are simultaneously wears by a user, one of them without battery management but still recording all the data, and the other with a smart battery algorithm. Battery discharge curves are shown and compared and a significant increase in the battery lifetime is demonstrated by using the algorithms.