Optimization of Electric-Vehicle Charging: scheduling and planning problems

The progressive shift from traditional vehicles to Electric Vehicles (EVs ) is considered one of the key measures to achieve the objective of a significant reduction in the emission of pollutants, especially in urban areas. EVs will be widely used in a not-so-futuristic vision, and new technologies will be present for charging stations, batteries, and vehicles. The number of EVs and Charging Stations (CSs) is increased in the last years, but, unfortunately, wide usage of EVs may cause technical problems to the electrical grid (i.e., instability due to intermittent distributed loads), inefficiencies in the charging process (i.e., lower power capacity and longer recharging times), long queues and bad use of CSs. Moreover, it is necessary to plan the CSs installation over the territory, the schedule of vehicles, and the optimal use of CSs. This thesis focuses on applying optimization methods and approaches to energy systems in which EVs are present, with specific reference to planning and scheduling decision problems. In particular, in smart grids, energy production, and storage systems are usually scheduled by an Energy Management System (EMS) to minimize costs, power losses, and CO2 emissions while satisfying energy demands. When CSs are connected to a smart grid, EVs served by CSs represent an additional load to the power system to be satisfied, and an additional storage system in the case of vehicle-to-grid (V2G) technology is enabled. However, the load generated by EVs is deferrable. It can be thought of as a process in which machines (CSs) serve customers/products (EVs) based on release time, due date, deadline, and energy request, as happens in manufacturing systems. In this thesis, first, attention is focused on defining a discrete-time optimization problem in which fossil fuel production plants, storage systems, and renewables are considered to satisfy the grid's electrical load. The discrete-time formalization can use forecasting for renewables and loads without data elaboration. On the other side, many decision variables are present, making the optimization problem hard to solve through commercial optimization tools. For this reason, an alternative method for the optimal schedule of EVs characterized by a discrete event formalization is presented. This new approach can diminish the number of variables by considering the time intervals as variables themselves. Of course, the solution's optimality is not guaranteed since some assumptions are necessary. Moreover, the last chapter proposes a novel approach for the optimal location and line assignment for electric bus charging stations. In particular, the model provides the siting and sizing of some CSs to maintain a minimum service frequency over public transportation lines.