In recent years, the considerable development of the transportation sector has made it the main contributor to energy consumption and greenhouse gas emissions. For example, transport accounts for 40% of CO2 emissions in France, where 82% of passengers travel by car. In addition to the climate crisis, the surge in international fossil fuel prices and the advancement in electric vehicle technology accelerated the adoption of electric vehicles as a great green alternative technology. According to the International Energy Agency, the number of electric vehicles attended 16.5 million in 2021, double the amount in 2019. Moreover, electric vehicle sales keep breaking records year after year. When electric vehicle adoption increases significantly, new challenges for electrical grid and charging infrastructure operators rise. On the one hand, the increasing power consumption due to charging will overload the grid and increase power losses and voltage deviation. On the other hand, charging infrastructure operators must meet the upcoming demands, maximize customer satisfaction, avoid long queuing, and minimize costs while respecting the power grid constraints. Therefore, these operators must adopt optimization strategies.
In this thesis, We consider a new charging station’s operating model where the charging station has limited total power and a limited number of chargers. Each charger is installed in a parking space. A reservation system is considered where electric vehicles submit their charging demands to avoid queuing. The scheduler allocates a suitable charger for each vehicle. Different objective functions were considered: minimizing the charging infrastructure capacity, maximizing the number of satisfied charging demands, maximizing the delivered energy, and minimizing total tardiness. In each case, the complexity of the problem is considered and proved in detail. We propose mathematical models, heuristics, and metaheuristics to jointly assign the electric vehicles to chargers and schedule the electric vehicle charging. Different comparison between various aspects of the charging schedule is given, namely: between identical and non-identical constant power rates, between the constant and variable charging power rates, between the preemptive and non-preemptive schedules, and between time-indexed and event-based models.