Probabilistic forecasting model for non-normally distributed EV charging demand

Daisuke Kodaira, Junji Kondoh

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

8 Citations (Scopus)

Abstract

A method for probabilistic electric vehicle (EV) demand forecasting is proposed in this paper. The EV demand in a certain area is forecasted by an ensemble forecasting model. The forecast result includes a deterministic forecasting and prediction interval that indicates the probability of deviation from deterministic forecasting. In the case study, an actual observed dataset from the UK is used to verify the proposed algorithm. The results show that the target EV charging demand to be forecasted shows 80% prediction interval coverage for 27 days out of 30 simulated days.

Original languageEnglish
Title of host publicationProceedings - 2020 International Conference on Smart Grids and Energy Systems, SGES 2020
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages623-626
Number of pages4
ISBN (Electronic)9781728185507
DOIs
Publication statusPublished - Nov 2020
Event2020 International Conference on Smart Grids and Energy Systems, SGES 2020 - Virtual, Perth, Australia
Duration: 23 Nov 202026 Nov 2020

Publication series

NameProceedings - 2020 International Conference on Smart Grids and Energy Systems, SGES 2020

Conference

Conference2020 International Conference on Smart Grids and Energy Systems, SGES 2020
Country/TerritoryAustralia
CityVirtual, Perth
Period23/11/2026/11/20

Keywords

  • Demand forecasting
  • Electric vehicle
  • Ensemble forecasting
  • Prediction interval
  • Probabilistic forecasting

Fingerprint

Dive into the research topics of 'Probabilistic forecasting model for non-normally distributed EV charging demand'. Together they form a unique fingerprint.

Cite this