Tampere University of Technology

TUTCRIS Research Portal

A hybrid optimization grey model based on segmented gra and multi-strategy contest for short-term power load forecasting

Research output: Contribution to journalArticleScientificpeer-review

Details

Original languageEnglish
Pages (from-to)15-28
Number of pages14
JournalJOURNAL OF GREY SYSTEM
Volume24
Issue number1
Publication statusPublished - 2012
Publication typeA1 Journal article-refereed

Abstract

In this paper, a hybrid grey model with both internal and external optimization is proposed to forecast the short-term power load which has the characteristics of nonlinear fluctuation and random growth. The internal optimization consists of modeling feasibility test and parameter a correction. The external optimization includes three aspects. First, the original series are selected from different viewpoints to construct different forecasting strategies. Second, the predicted day is divided into several smooth segments for separate forecasting. Finally, the different forecasting strategies are implemented respectively in the different segments through grey correlation contest. A practical application verifies that the proposed model has a higher forecasting accuracy and the independency on the choice of initial value.

Keywords

  • External optimization, Hybrid grey model, Multi-strategy contest, Parameter a correction, Segmented grey correlation, Short-term power load forecasting