The availability of charging infrastructure is essential for large-scale adoption of electric vehicles (EV). Charging patterns and the utilization of infrastructure have consequences not only for the energy demand by loading local power grids, but influence the economic returns, parking policies and further adoption of EVs. We develop a data-driven approach that exploits predictors compiled from Geographic Information Systems data describing the urban context and urban activities near charging infrastructure to explore correlations with a comprehensive set of indicators that measure the performance of charging infrastructure. The best fit was identified for the size of the unique group of visitors (popularity) attracted by the charging infrastructure. Consecutively, charging infrastructure is ranked by popularity. The question of whether or not a given charging spot belongs to the top tier is posed as a binary classification problem and predictive performance of logistic regression regularized with an l1 penalty, random forests and gradient boosted regression trees is evaluated. Obtained results indicate that the collected predictors contain information that can be used to predict the popularity of charging infrastructure. The significance of predictors and how they are linked with the popularity are explored as well. The proposed methodology can be used to inform charging infrastructure deployment strategies.

Transport systems are expected to widely shift towards electric propulsion in the next decade. The diffusion of Electrical Vehicles (EVs) however creates great challenges; among the others, EV charging patterns are non-controllable, thus it is mandatory to have at disposal high quality EV load forecasts in order to reach the operational excellence of networks with a wide EV penetration. Relevant literature on EV load forecasting is quite scarce, compared to other load forecasting applications; this paper aims at filling this gap by providing a comparative study between the performances of time series and machine learning approaches. The comparative analysis is performed on actual EV load data, extracted from a dataset collected at 1700 charging stations in the Netherlands. The results of numerical experiments are given in terms of aggregate energy consumption for lead times up to 28 days ahead, in order to fully suit the time horizons typical of distribution systems management.

The rapid adoption of electric vehicles (EVs) has catalyzed the urgency of a resilient and sustainable charging infrastructure. This study addresses this issue by harnessing offshore wind energy resources to meet the charging demands, particularly in remote coastal regions. The primary challenge tackled in this research involves the complex management of power flow dynamics due to the inherent variability of wind energy and the stochastic nature of EV charging demands, modeled using probabilistic distributions to represent varying arrival times, charging durations, and power requirements of EVs. This work introduces a pioneering framework centered on the development of a wind-powered electric vehicle charging system (EVCS) that utilizes a medium-voltage direct current (MVDC) bus. An enhanced decentralized model predictive control (MPC) strategy was employed that distinguishes itself from conventional control paradigms due to its heightened adaptability and proficient management of the dynamic interactions among wind energy generation, energy storage systems (ESS), EV charging demands, and grid interactions. Rigorous simulations and real-time hardware-in-loop studies underscore the efficacy of the MPC strategy in preserving the voltage stability within the MVDC bus while optimizing the power flow, thereby minimizing energy losses and ensuring grid resilience. These results validate the viability of the proposed wind energy-integrated EVCS as an integral component of seaport grid infrastructure.

Here, we develop a data-centric approach to analyse which activities, functions, and characteristics of the environment surrounding the slow charging infrastructure impact the distribution of the electricity consumed at slow charging infrastructure. We analysed the probability distribution of energy consumption and its relation to indicators characterising charging events to gain basic insights. The energy consumption can be satisfactorily modelled by a transformed beta distribution and the number of charging transactions is the driving factor among the characteristics constituting the energy consumption. We collected geospatial datasets and prepared a large number of candidate features modelling the spatial context in which the charging infrastructure operates. Using statistical methods, we identified and interpreted a relatively small subset of the most influential features correlated with energy consumption. The majority of these features are related to the economic prosperity of residents. Residents and businesses with high (low) income, situated nearby charging infrastructure, are linked to a positive (negative) impact on energy consumption. Similarly, charging infrastructure located close to expensive newly built housing shows higher energy consumption. The largest adverse impact has the high concentration of residents receiving social assistance. By applying the methodology to a specific charging infrastructure class, e.g. determined by the used rollout strategy, we differentiated the selected features. Business types, working sector of residents and public venues in the proximity are linked to higher consumption of energy at charging infrastructure deployed strategically. Characteristics linked with the age structure of the population are linked to the energy consumption at charging infrastructure placed based on the demand. Data collection and data processing are among the most time-consuming activities. The paper provides valuable insights into which data to collect and use as features when developing prediction models to inform charging infrastructure deployment and planning of power grids.