Towards Accurate Modeling of Public EV Charging Loads for Efficient Charger Network Expansion

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21230%2F24%3A00375070" target="_blank" >RIV/68407700:21230/24:00375070 - isvavai.cz</a>

Výsledek na webu

<a href="https://datascience.stanford.edu/2024-stanford-data-science-conference-poster-session-1" target="_blank" >https://datascience.stanford.edu/2024-stanford-data-science-conference-poster-session-1</a>

DOI - Digital Object Identifier

—

Jazyk výsledku

angličtina

Název v původním jazyce

Towards Accurate Modeling of Public EV Charging Loads for Efficient Charger Network Expansion

Popis výsledku v původním jazyce

Just like in other parts of the world, the transition towards electric vehicles (EVs) in Europe, driven by the European Union's carbon reduction targets, underscores the critical need for strategic planning of public charging infrastructure expansion. However, while adoption of EVs is growing exponentially, buildup of charging points is slowing down as most favourable locations get depleted. In our contribution we propose a method for optimizing the development of charging infrastructure by integrating a machine learning based model incorporating observed charging behavior and electromobility growth scenarios. Based on real-world data collected from public charging stations, the research estimates EV charging demand under various scenarios for 2030 and 2050. Additionally, geospatial analysis categorizes potential charging station locations into residential, commercial, and leisure areas to guide strategic deployment. The study employs a three-step machine learning approach, encompassing data analysis, generative modelling, and demand projection, to provide insights into future charging infrastructure needs. Results suggest that accurate demand forecasting can facilitate efficient allocation of resources and support the decarbonization of the transportation sector. Furthermore, the study's methodology and findings can be replicated in other cities, offering a valuable tool for stakeholders and infrastructure developers in planning the strategic deployment of charging stations.

Název v anglickém jazyce

Towards Accurate Modeling of Public EV Charging Loads for Efficient Charger Network Expansion

Popis výsledku anglicky

Just like in other parts of the world, the transition towards electric vehicles (EVs) in Europe, driven by the European Union's carbon reduction targets, underscores the critical need for strategic planning of public charging infrastructure expansion. However, while adoption of EVs is growing exponentially, buildup of charging points is slowing down as most favourable locations get depleted. In our contribution we propose a method for optimizing the development of charging infrastructure by integrating a machine learning based model incorporating observed charging behavior and electromobility growth scenarios. Based on real-world data collected from public charging stations, the research estimates EV charging demand under various scenarios for 2030 and 2050. Additionally, geospatial analysis categorizes potential charging station locations into residential, commercial, and leisure areas to guide strategic deployment. The study employs a three-step machine learning approach, encompassing data analysis, generative modelling, and demand projection, to provide insights into future charging infrastructure needs. Results suggest that accurate demand forecasting can facilitate efficient allocation of resources and support the decarbonization of the transportation sector. Furthermore, the study's methodology and findings can be replicated in other cities, offering a valuable tool for stakeholders and infrastructure developers in planning the strategic deployment of charging stations.

Druh

O - Ostatní výsledky

CEP obor

—

OECD FORD obor

20205 - Automation and control systems

Projekt

—

Návaznosti

S - Specificky vyzkum na vysokych skolach

Rok uplatnění

2024

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů