Skip to main navigation Skip to search Skip to main content

Short-term load forecasting for the electric bus station based on GRA-DE-SVR

  • North China Electric Power University

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

19 Scopus citations

Abstract

With large-scale electric vehicles penetrating into power system, the grid will be faced with severe challenges. Accurate charging load forecasting is required to ensure the security and economy of the grid. Firstly, the factors that influence the daily load of electric bus stations are analyzed in this paper. Based on the grey relation theory, samples of similar days are selected to establish SVM prediction model. In order to improve prediction accuracy, differential evolution (DE) algorithm is applied to optimize parameters of SVR model. Through empirical study, the root mean square error (RMSE) of daily load forecasting is 10.85%. Compared with the standard SVM prediction model, the prediction precision of this paper is increased by 1.52%. What's more, the proposed method has better forecasting performance than the other methods.

Original languageEnglish
Title of host publication2014 IEEE Innovative Smart Grid Technologies - Asia, ISGT ASIA 2014
PublisherIEEE Computer Society
Pages388-393
Number of pages6
ISBN (Print)9781479913008
DOIs
StatePublished - 2014
Externally publishedYes
Event2014 IEEE Innovative Smart Grid Technologies - Asia, ISGT Asia 2014 - Kuala Lumpur, Malaysia
Duration: 20 May 201423 May 2014

Publication series

Name2014 IEEE Innovative Smart Grid Technologies - Asia, ISGT ASIA 2014

Conference

Conference2014 IEEE Innovative Smart Grid Technologies - Asia, ISGT Asia 2014
Country/TerritoryMalaysia
CityKuala Lumpur
Period20/05/1423/05/14

Keywords

  • Electric vehicles
  • differential evolution
  • grey relation analysis
  • short-term load forecasting
  • support vector machine

Fingerprint

Dive into the research topics of 'Short-term load forecasting for the electric bus station based on GRA-DE-SVR'. Together they form a unique fingerprint.

Cite this