Predictive energy management in an electric vehicle charging station - Université de Picardie Jules Verne Accéder directement au contenu
Communication Dans Un Congrès Année : 2022

Predictive energy management in an electric vehicle charging station

Résumé

This paper describes an energy management system (EMS) based on photovoltaic (PV) production forecasts to optimize energy flows in a microgrid. The microgrid is composed of 6 electric vehicles (EV), battery, and a PV production system. The goal is to maximize EV charging while considering moments of PV non-production. The proposed EMS is composed of two elements: a deep-learning model using LSTM to forecast PV production and a rule-based algorithm to dispatch power flow in the microgrid. The simulation findings reveal that the proposed power management system is capable of greatly increasing the state of charge (SOC) of EVs while anticipating times of PV non-generation. Four scenarios were run over different prediction periods to show the benefit of forecasting on energy management. A total benefit of 9.49% on the total state of charge (SOC) of EVs charged with a 60-minute forecast compared to management without PV forecast.
Fichier non déposé

Dates et versions

hal-04264407 , version 1 (30-10-2023)

Identifiants

Citer

M. Kermia, Jérôme Bosche, D. Abbes. Predictive energy management in an electric vehicle charging station. CIRED Porto Workshop 2022: E-mobility and power distribution systems, Jun 2022, Porto, Portugal. pp.935-939, ⟨10.1049/icp.2022.0851⟩. ⟨hal-04264407⟩

Collections

U-PICARDIE MIS
22 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More