Prediction interval estimation of demand curve in electric power distribution system

Tomohide Yamazaki, Shinji Wakao*, Hirokazu Ito, Tsuneyo Sano

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review


In a distribution line, power system control and power equipment investment are planned based on a measured power system current. However, recently the mass introduction of PV make it difficult for us to precisely measure the demand curve that is a current consumed by electrical equipment because the reversal power flow from PV systems is superposed. Therefore, the prediction of demand curves of distribution line is indispensable for power system management. Additionally, it is also necessary to estimate the reliability of the predicted values as well as predicted current itself. In this paper, we propose the estimation method of the prediction interval that is the index of reliability based on the past demand curve database. The feature of the proposed method based on Just-In-Time (JIT) modeling make it possible for us to accurately estimate the prediction interval by the normalized database of demand curve. In this paper, some numerical examples are presented, which demonstrate the effectiveness of the proposed method.

Original languageEnglish
Pages (from-to)848-857
Number of pages10
JournalIEEJ Transactions on Power and Energy
Issue number12
Publication statusPublished - 2016


  • Cluster analysis
  • Electric power distribution system
  • Just-in-time modeling
  • Kernel density estimation
  • Load forecasting
  • Prediction interval

ASJC Scopus subject areas

  • Energy Engineering and Power Technology
  • Electrical and Electronic Engineering


Dive into the research topics of 'Prediction interval estimation of demand curve in electric power distribution system'. Together they form a unique fingerprint.

Cite this