Medium term load forecasting in distribution systems based on multi linear regression & principal component analysis: A novel approach: A novel approach

Roozbeh Torkzadeh, Ahmad Mirzaei, Mohammad Mehdi Mirjalili, Alireza Sedighi Anaraki, Mohammad Reza Sehhati, Farideh Behdad

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

27 Citations (Scopus)

Abstract

An accurate medium term load forecast (MTLF) is essential for expansion planning studies of distribution systems. Also the mid-term electric load as a function of time has a complex nonlinear behavior which makes the ordinary linear prediction methods seems insufficient. In this paper, a combination of multi linear regression and principle components analysis is used to predict weekly electrical peak load of Yazd city distribution system. According to the prediction results, main benefits of proposed method are simplicity of calculations and high accuracy forecasting for multi-horizon predictions. MATLAB© is used to implement the forecaster model.
Original languageEnglish
Title of host publication2014 19th Conference on Electrical Power Distribution Networks, EPDC 2014
Place of PublicationPiscataway
PublisherInstitute of Electrical and Electronics Engineers
Pages66-70
Number of pages5
ISBN (Electronic)9781479956364
DOIs
Publication statusPublished - 6 May 2014
Event2014 19th Conference on Electrical Power Distribution Networks, EPDC 2014 - Tehran, Iran, Islamic Republic of
Duration: 6 May 20147 May 2014

Conference

Conference2014 19th Conference on Electrical Power Distribution Networks, EPDC 2014
Country/TerritoryIran, Islamic Republic of
CityTehran
Period6/05/147/05/14

Keywords

  • correlation analysis
  • mid-term load forecast
  • multi linear regression
  • principel components analysis

Fingerprint

Dive into the research topics of 'Medium term load forecasting in distribution systems based on multi linear regression & principal component analysis: A novel approach: A novel approach'. Together they form a unique fingerprint.

Cite this