In the passing decades, the monitoring of pavements and passing vehicles was developed vigorously with the growth of information and sensing technology. Pavement monitoring is an essential part of pavement research and plays an important role in transportation system. At the same time, the monitoring system about the traffic, such as Weigh-in-Motion (WIM) system and traffic classification system, also attracted lots of attention because of their importance in traffic statistics and management. The monitoring system in this dissertation combines the monitoring for pavements and traffic together with the same sensing network. For pavement health monitoring purpose, the modulus of the asphalt layer can be back-calculated based on the collected mechanical responses under corresponding environmental conditions. At the same time, the actually strain and stress in pavements induced by each passing vehicle are also used for pavement distress prediction. For traffic monitoring purpose, the horizontal strain traces are analyzed with a Gaussian model to estimate the speed, wandering position, weight and classification of each passing vehicle. The whole system, including the sensing network and corresponding analysis method, can monitor the pavement and the traffic simultaneously, and is called transportation monitoring system. This system has a high efficiency because of its low cost and easy installation; multi-functionality to provide many important information of transportation system. Many related studies were made to improve the prototyped transportation monitoring system. With the assistance of numerical simulation software ABAQUS and 3D-Move, the effect of many loading and environmental conditions, including temperature, vehicle speed, tire configuration and inflation pressure, are taken into consideration. A method was set up to integrate data points from many tests of similar environmental and loading conditions based on Gaussian model. Another method for consistent comparison of variable field sensor data was developed. It was demonstrated that variation in field measurement was due to uncontrollable environmental and loading factors, which may be accounted for by using laboratory test and numerical simulation based corrections. / Ph. D.

L’évaluation des caractéristiques structurelles des chaussées permet l’optimisation de leur maintenance. La mesure de déflexion est une mesure de base de cette évaluation. Aujourd’hui,seuls la déflexion maximale et le rayon de courbures ont analysés. Pourtant, le curviamètre et le déflectographe relèvent le bassin de déflexion complet dont l’analyse permettrait d’extraire des paramètres plus sensibles à l’endommagement des chaussées. Actuellement, l’interprétation des mesures est seulement qualitative et aucun calcul inverse de l’état de dommage de la structure n’est réalisé. La thèse vise à améliorer l’interprétation des mesures de déflexion. Le fonctionnement des appareils a montré que leurs hypothèses de mesure conduisent à des biais de mesure. Pour pallier ce problème, une procédure de correction des mesures a été mise en place. La comparaison avec des bassins théoriques a montré que la méthode de correction était satisfaisante. Ensuite, une étude numérique visant à déterminer la sensibilité du bassin de déflexion à la présence de différents défauts a été conduite. Celle ci a montré que les indicateurs classiques de la mesure de déflexion étaient peu sensibles à un endommagement dans la chaussée. Une méthodologie a donc été développée pour la création d’indicateurs optimisés à un type de défauts spécifiques. L’étude sur cas théoriques a conduit à l’obtention de résultats concluants. L’ensemble de ces travaux a ensuite été validé sur sites expérimentaux pour des mesures de répétabilité, mais également sur site avec la présence avérée de défauts. Finalement, la thèse a envisagé l’utilisation des travaux précédents sur des mesures réelles recueillies sur itinéraire. / The evaluation of the structural characteristics of pavements is involved in their maintenance. The measurement of deflection is a key indicator of this evaluation. Currently, only the maximum deflection and the radius of curvature are analyzed. However, the curviameter and the deflectograph are able to record the whole deflection bowl and the parameters deduced from thismeasurement could help to better characterized damages on pavements. The interpretation of the measurements is only qualitative and back calculation of pavement layer moduli gives unsatisfactory results. The thesis aims to improve the interpretation of deflection measurements. The working principle of these apparatus and the measurement assumptions introduce several measurement biases.To overcome these issues, a correction process has been developed. The comparison with theoretical basins has given satisfactory results on bituminous or flexible pavements. Then, a numerical study has been conducted to determine the influence of pavements damages on the deflection measurement. By this study, it has been showed that the usual indicators of the deflection measurement are notable to detect all damages. So, a methodology hasbeen developed in order to create an optimized indicator specified to a special defect. A study on numerical results has been conducted to validate the implementation of these indicators. Then, the correction method and these new indicators have been tested on experimental sites with and without damages. At last, the research works have been studied at network level.