Movatterモバイル変換

evprof

Overview

evprof is part of a suite of packages to analyse, model and simulatethe charging behavior of electric vehicle users:

• evprof: ElectricVehicle PROFiling
• evsim: ElectricVehicle SIMulation

evprof aims to provide tools for classifying EV charging sessionsinto generic groups with similar connection patterns named “userprofiles”, using the Gaussian Mixture Models (GMM) clustering method.Moreover, functions to build stochastic models (based on GMM) for everyuser profile are also provided in order to simulate new EV sessions.

The Gaussian Mixture Models clustering technique used in this packageaims to accomplish two different tasks that can be useful for multiplepurposes:

1. Classification of EV charging sessions into generic user profiles(e.g. working time, dinner, commuters, etc.), allowing to:

• Increase the knowledge on the different flexibility potentialpatterns from a real data set
• Define accurate tariffs according to the flexibility potential(implicit demand response scenario)
• Reduce the uncertainty of flexibility offers when participating inflexibility markets (explicit demand response scenario)

2. Modeling every user profile with stochastic models, allowingto:

• Simulate high penetration of EV to estimate when an existingcharging infrastructure will be saturated
• Simulate different scenarios of charging rates to analyse the impactof fast charging
• Size and plan a public charging infrastructure

Usage

To use this package you will need a data set of EV charging sessionswith at least two fundamental variables:connectionstart time andconnection duration. With thesetwo variables you will be able to classify the sessions into differentuser profiles, but to generate the EV Gaussian Models you will also needtheenergy values.

The package also provides an example open data set of EV chargingsessions from the California Technological Institute (Caltech), whichcan be downloaded from theACN-Data website. For moreinformation about this data set and how to use it, visit theACNdocumentation. Moreover, an exampleevmodel object (EVGaussian Mixture Models) built withevprof functions andthe California open data set (see theCaliforniacase study article) is also provided. These two demo data objectsare provided together with package functions for a better interactiveuser experience.

If you have your own data set, the best place to start is theGetstarted chapter in the package website.

Installation

You can install the package from CRAN or the development version fromGitHub:

# CRAN stable releaseinstall.packages("evprof")# Latest development version# install.packages("devtools")devtools::install_github("mcanigueral/evprof")

Getting help

If you encounter a clear bug, please open an issue with a minimalreproducible example onGitHub. Forquestions and other discussion, please send me a mail tomarc.canigueral@udg.edu.

For further technical details, you can read the following academicarticles about the methodology used in this paper:

• Electric vehicle user profiles for aggregated flexibilityplanning. IEEE PES Innovative Smart Grid Technologies Europe(ISGT Europe). IEEE, Oct. 18, 2021.DOIlink.
• Flexibility management of electric vehicles based on userprofiles: The Arnhem case study. International Journal ofElectrical Power and Energy Systems, vol. 133. Elsevier BV, p. 107195,Dec. 2021.DOIlink.
• Potential benefits of scheduling electric vehicle sessionsover limiting charging power. CIRED Porto Workshop 2022:E-mobility and power distribution systems. Institution of Engineeringand Technology, 2022.DOIlink.
• Assessment of electric vehicle charging hub based onstochastic models of user profiles. Expert Systems withApplications (Vol. 227, p. 120318). Elsevier BV. May 2023.DOI link.

Acknowledgements

This work has been developed under a PhD program in theeXiT research group from the Universityof Girona (Catalonia) in collaboration withResourcefully, an energy transitionconsulting company based in Amsterdam, The Netherlands.