Optimisation of the deployment of electric vehicle chargers

Abstract

This thesis introduces a new model based on Markov chains that provides both single and multiperiod versions, allowing the modelling of user behavior within a fixed time frame (e.g., an hour) as well as accommodating user behavior across a time span where their behavior varies over time (e.g., a 24-hour period) and will mainly focus on the formulation of this problem, the formulation of transition matrices associated to the model (transition matrices with probabilities of going from one to another location and use of battery states), and how to obtain them through real-world data (with the use of graphs, clustering and shortest path algorithms). The work also presents a model from the literature that addresses the problem, and a comparison will be established. A section will also be dedicated to generating artificial data (referred to as ’Approximated data’) in order to study the behavior of the model through a variety of tests. A section will also be dedicated to the mobility data that can be found in Belgium and how to use them for the Markov model.