Robust data-driven models and algorithms for the electric vehicle ecosystem

SUMMARY OF THE RESEARCH PROPOSAL

The aim is to develop decision-making tools for the management of the electric vehicle ecosystem. The first step is to develop efficient and robust machine learning models for predicting the energy consumption of electric vehicles in different scenarios and taking onto account a variety of endogenous and exogenous variables (traffic, weather, driving behavior, etc.). Then, using energy consumption models, methods to optimize the routing of vehicles to charging stations will be developed. Further, energy management and scheduling algorithms at grid-tied charging stations will be developed in different scenarios (with/without renewable energy sources, with/without storage capability). The research methodology is based on combinatorial optimization and machine learning. The approach will be tested using real electric vehicles and a real PV-backed charging station. 

REQUIRED ACADEMIC QUALIFICATIONS & SKILLS

• Master degree or engineering degree in Operations Research/Data Science/computational Mathematics or equivalent
• Strong knowledge in operations research and/or machine learning.
• Excellent programming skills (e.g., C/C++, Python, or Java)
• Knowledge of optimization software (CPLEX, GUROBI, Local solver, GAMS, etc.) is a plus
• Knowledge of Machine Learning software (Scikit-learn, PyTorch, TensorFlow, etc.) is a plus
• Strong written and verbal communication skills in English

The PhD student will spend 18 months at University of Lorraine and 18 months at Université Internationale de Rabat - UIR

HOW TO APPLY

• Send a letter of motivation, transcript of grades, and your CV to:    ammar.oulamara@loria.fr     and  mounir.ghogho@uir.ac.ma

Host research units: LORIA – Université de Lorraine and TICLab  - College of Engineering and Architecture - Université Internationale de Rabat - UIR

Supervisors: Ammar Oulamara (Université de Lorraine), Mounir Ghogho (UIR), Mustapha Oudani (UIR)