Thesis defense of Abdelgader ABDELMUHSEN
Abdelgader ABDELMUHSEN, PhD student at Lames, Université Gustave Eiffel, has defended his thesis entitled
"Application of Artificial Intelligence in the exploitation of deflection measurements for the mechanical characterization of pavement structures"
on
monday september 30th 2024
at 9.30 AM at the amphitheatre of Université Gustave Eiffel, campus of Nantes (France)
The thesis was presented before the following jury :
Reporters :
Prof. Simona Fontul, Assistant Professor - LNEC, Lisbonne-Portugal
Prof. Khalil EL Khamlichi Drissi, Professeur des universités - Université Clermont Auvergne
Examiners :
Prof. Fekri Meftah, Professeur des universités - INSA Rennes
Prof. Yide Wang, Professeur des universités - Université de Nantes
Thesis directors :
Dr. Amine Ihamouten, IDTPE - HDR - Université Gustave Eiffel
Dr. Franziska Schmidt, IDTPE - HDR - Université Gustave Eiffel
Supervisor :
Dr. Jean-Michel Simonin, Chargé de recherche - Université Gustave Eiffel
Abstract :
This thesis presents an effective model and methodology based on machine learning (ML) for estimating the elastic modulus of pavement layers (Ei). This innovative approach uses deflection slope measurements from TSD (DS) rather than the classical deflection method. The inverse model developed in this research is based on data from industry-recognized software (Alizé-LCPC) and an external experimental simulator, ensuring their conformity with real-world conditions. An advanced parametric validation process was also executed to validate the model performance. Moreover, a quantitative sensitivity analysis has been conducted to investigate the negative impacts of four major challenges: the complexity of highdimensional TSD data, the influence of measurement noise, temperature variations, and uncertainties in Ei of other layers.
All of this aims to underscore the primary objective: to make valuable contributions to the field of pavement mechanics.
Keywords : Bearing Capacity Assessment, Pavement Behavior Modeling, Pavement Elastic Modulus, Traffic Speed Deflectometer, Deflection Slope, Machine Learning