Retour au site du LAAS-CNRS

Abstract

The work detailed in this document is the result of a collaborative effort of the LAASCNRS in Toulouse and MEAS-France / TE Connectivity during a period of three years. The goal here is to develop a methodology to design smart embedded sensors with the ability to estimate physical parameters based on multi-physical data fusion. This strategy tends to integrate sensors technologies, currently dedicated to lab measurements, in low powered embedded systems working in imperfects environments. After exploring model oriented methods, parameters estimations and Kalman filters, we detail various existing solutions upon which we can build a valid response to multiphysical data fusion problematics, for linear systems with the Kalman Filter, and for non-linear systems with the Extended Kalman Filter and the Unscented Kalman Filter. Then, we will synthesize a filter for hybrid systems, having an linear evolution model and a non-linear measurement model. For example, using the best of the two worlds in order to obtain the best complexity/precision ratio. Once we selected the estimation method, we detail computing power and algorithm complexity problematics in order to find available optimizations we can use to assess the usability of our system in a low power environment. Then we present the developed methodology application to the UQS sensor, sold by TE Connectivity, study case. This sensor uses near infra-red spectroscopy to determine the urea concentration in a urea/water solution, in order to control the nitrogen-oxyde depolluting process in gasoline engines. After a design principles presentation, we detail the model we created in order to represent the system, to simulate its behavior and to combine the measurement data to extract the desired concentration. During this step, we focus on the obstacles of our model calibration and the deviation compensation, due to working conditions or to components aging process. Based on this development, we finally designed the hybrid models addressing the nominal working cases and the model re-calibration during the working duration of the product. After this, we presented obtained results, on simulated data, and on real-world measured data. Finally, we enhanced the methodology based on tabulated “black box” models which are easier to calibrate and cheaper to process. In conclusion, we reapplied our methodology to a different motion capture sensor, to compile all possible solutions and limits.

Résumé

Les travaux de thèse présentés ici sont le fruit d’un travail collaboratif effectué entre le LAAS-CNRS de Toulouse et l’entreprise MEAS-France / TE Connectivity sur une durée de trois ans. L’objectif est ici de mettre en place une méthodologie permettant la réalisation de capteurs embarqués intelligents capables d’estimer des paramètres physiques à partir de la fusion de données multi-physiques, cette stratégie devant permettre l’intégration de capteurs, jusqu’alors réservés aux mesures en laboratoire dans des environnement défavorables, au sein de systèmes embarqués disposant de faibles puissances de calculs. Après une exploration des méthodes liées à la modélisation, l’estimation de paramètres et les filtres de Kalman, nous détaillons différentes solutions existantes sur lesquelles nous appuyer pour répondre à la problématique de la fusion de données multi-physiques, pour les systèmes linéaires, avec le filtre de Kalman, les systèmes non-linéaires avec les filtres de Extended Kalman Filter, et Unscented Kalman Filter. Puis nous synthétisons un filtre pour les systèmes hybrides, présentant par exemple une évolution linéaire et une mesure non-linéaire, utilisant le meilleur des deux parties afin de tendre vers le meilleur compromis complexité/précision. Une fois la méthode d’estimation sélectionnée, nous nous attardons sur les problématiques liées à la puissance de calcul et à la complexité algorithmique, pour chercher les optimisations pouvant êtres mises en place afin de garantir le fonctionnement sur un système embarqué disposant d’une faible puissance de calcul. Nous exposons ensuite l’application de cette méthodologie sur le cas d’étude du capteur UQS, commercialisé par TE Connectivity, utilisant la spectroscopie proche infrarouge pour déterminer la concentration d’urée en solution dans le cadre de la régulation des systèmes de dépollution des oxydes d’azote sur les moteurs diesel. Après une présentation des principes de fonctionnement, nous présentons le modèle développé que nous avons dressé afin de représenter le système, de simuler son comportement et d’effectuer la fusion de mesures permettant d’extraire la concentration désirée. Au cours de cette étape, nous abordons les problématiques liées à la calibration du modèle, ainsi qu’à la compensation de dérives, qu’elles soient liées aux conditions d’utilisation ou au vieillissement des composants. En s’appuyant sur cette réflexion, nous dressons finalement les modèles hybrides nous permettant d’aborder les cas de fonctionnement nominaux, ainsi que les recalibrations du modèle au cours de la vie du produit. Après quoi, nous présentons les résultats obtenus, d’abord sur des données simulées, puis sur des données capturées sur un système réel, avant de proposer des pistes d’amélioration de la méthodologie en se basant sur des modèles tabulés en « boîte noire » qui permettent de faciliter l’étape de calibration tout en diminuant la charge de calcul. Finalement, nous généralisons la méthodologie par le biais d’une problématique de capture de mouvement, pour conclure sur les possibilités et les limites de la solution que nous avons proposée.

Mots-Clés / Keywords

Capteurs intelligents; Filtrage de Kalman; Fusion de données; Systèmes embarqués; Smart sensing; Kalman filter; Data fusion; Embedded systems;

Abstract

Transdermal administration might be an interesting alternative to current routes of administration. It is noninvasive, avoids gastric side effects, and improves bioavailability of the molecules. However, due to the low permeability of the stratum corneum, a permeation enhancement strategy is required to make a large number of molecules suitable to this mode of administration. To overcome those limitations and deliver controlled drugs, a smart transdermal drug delivery system is designing. This paper focusses on the study of transdermal iontophoresis permeation of lidocaine hydrochloride (2%, w/w) across pig ear skin. In vitro iontophoresis experiments were conducted using Franz diffusion cells. Anodal iontophoresis was applied for 30 minutes at different current densities. Samples (1 mL) were withdrawn every 30 minutes from the receptor compartment and replaced with a fresh buffer and then analyzed using High Performance Liquid Chromatography (HPLC). Results highlight the relationship between current density, time of stimulation, and amount of lidocaine permeated.

Abstract

The objective of this study is to determine the effect of iontophoresis on the transport of lidocaine hydrochloride (2%, w/w) across ear pig skin. Our project consists of the development of a transdermal drug delivery system (TDDS). The system is designed as a reusable electronic patch where micropump, reservoir, sensors and Bluetooth communication are integrated. Therefore, quantify the iontophoretic diffusion and investigate the effect of different current densities in enhancing and controlling drug delivery is a primordial step. In vitro iontophoresis experiments were conducted using Franz diffusion cells. Iontophoresis was applied for 30 minutes at current density of 0.25, 0.5 and 0.75 mA/cm2. Samples (1 mL) were withdrawn every 30 minutes from the receptor compartment and replaced with fresh buffer and then analyzed using High Performance Liquid Chromatography (HPLC). The amount of lidocaine HCl delivered was 247 ± 19, 199 ± 14 and 115 ± 7.11 μg/cm2 for 0.75, 0.5 and 0.25 mA/cm2 respectively and 25.6 ± 3.02 μg/cm2 for passive diffusion. The amount is influenced by current density, and could be therefore enhanced, predicted and controlled by those parameters.

Abstract

The present work aimed to assess the validity and reliability of the 3D motion analyzer (Shimano Dynamics Lab, Sittard, Netherland) during laboratory cycling tests in comparison with the Vicon device (Vicon Motion Systems Ltd. Oxford, UK). Three cyclists were required to complete one laboratory cycling test at three different pedalling cadence and at a constant power output. Kinematic measurements were collected simultaneously from 3D motion analyzer and Vicon devices and performed five times for each pedalling cadence. The two systems showed a high reliability with excellent intraclass correlation coefficients for most kinematic variables. Moreover, this system was considered as valid by considering the error due to the initial markers placement. Experts and scientists should use the Vicon system for the purpose of research whereas the 3D motion analyzer could be used for bike fitting.

Abstract

The present work aimed to assess the effective frontal area (AC d , m 2) of a cyclist using both 3D scanning and Computational Fluid Dynamics (CFD) simulation and compare the results with wind tunnel and field measurements. One elite cyclist was recruited to complete a 3D scanning, a wind tunnel test and a field test. The 3D scanning was analyzed using CFD simulation to determine the AC d of the cyclist. The CFD AC d was compared to those measured in both wind tunnel and field tests. The 3D scanning method provides useful data for cycling science and TT position or equipment optimization, by using iterative approach. Indeed, the AC d obtained after CFD simulation was in accordance with those obtained in both wind tunnel and field testing sessions. Resolution, scanning time and post processing are compatible with an extensive use in real conditions and with a larger number of cyclists.

Abstract

This article presents a multimodal observation method of older people discovering the use of a smart house and tactile and spoken interaction modes inside a Living Lab. Our observation is based on ethnographic and numerical methods. We highlight the impact of presentation of the different technologies on the change of mind, the preference for voice interaction and some brakes considered as far as the adoption of new technologies is concerned.