A Gaussian Processes Technique for Short-term Load Forecasting with Considerations of Uncertainty
-
- Ohmi Masataro
- Dept. of Electrical and Electronics Eng., Meiji University
-
- Mori Hiroyuki
- Dept. of Electrical and Electronics Eng., Meiji University
Bibliographic Information
- Other Title
-
- ガウシアンプロセスによる不確定性を表現した短期電力負荷予測
- ガウシアンプロセス ニ ヨル フカクテイセイ オ ヒョウゲン シタ タンキ デンリョク フカ ヨソク
Search this article
Abstract
In this paper, an efficient method is proposed to deal with short-term load forecasting with the Gaussian Processes. Short-term load forecasting plays a key role to smooth power system operation such as economic load dispatching, unit commitment, etc. Recently, the deregulated and competitive power market increases the degree of uncertainty. As a result, it is more important to obtain better prediction results to save the cost. One of the most important aspects is that power system operator needs the upper and lower bounds of the predicted load to deal with the uncertainty while they require more accurate predicted values. The proposed method is based on the Bayes model in which output is expressed in a distribution rather than a point. To realize the model efficiently, this paper proposes the Gaussian Processes that consists of the Bayes linear model and kernel machine to obtain the distribution of the predicted value. The proposed method is successively applied to real data of daily maximum load forecasting.
Journal
-
- IEEJ Transactions on Power and Energy
-
IEEJ Transactions on Power and Energy 126 (2), 202-208, 2006
The Institute of Electrical Engineers of Japan
- Tweet
Keywords
Details 詳細情報について
-
- CRID
- 1390282679578254208
-
- NII Article ID
- 10017153883
- 30011601517
-
- NII Book ID
- AN10136334
-
- BIBCODE
- 2006IJTPE.126..202O
-
- ISSN
- 13488147
- 03854213
-
- NDL BIB ID
- 7818923
-
- Text Lang
- ja
-
- Data Source
-
- JaLC
- NDL
- Crossref
- CiNii Articles
- KAKEN
-
- Abstract License Flag
- Disallowed