Empirical Mode Decomposition based Multi-objective Deep Belief Network for short-term power load forecasting

With the rapid development of power grid data, the data generated by the operation of the power system is increasingly complex, and the amount of data increases exponentially. In order to fully exploit and utilize the deep relationship between data to achieve accurate prediction of power load, this...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Neurocomputing (Amsterdam) Jg. 388; S. 110 - 123
Hauptverfasser: Fan, Chaodong, Ding, Changkun, Zheng, Jinhua, Xiao, Leyi, Ai, Zhaoyang
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Elsevier B.V 07.05.2020
Schlagworte:
Deep belief network
Empirical Mode Decomposition
Ensemble learning
Multi-objective optimization algorithm
Power load forecasting
Empirical Mode Decomposition
Multi-objective optimization algorithm
Power load forecasting
Ensemble learning
Deep belief network
ISSN:0925-2312, 1872-8286
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:With the rapid development of power grid data, the data generated by the operation of the power system is increasingly complex, and the amount of data increases exponentially. In order to fully exploit and utilize the deep relationship between data to achieve accurate prediction of power load, this paper proposes an Empirical Mode Decomposition Based Multi-objective Deep Belief Network prediction method (EMD-MODBN). In the training process of DBN, a multi-objective optimization model is constructed aiming at accuracy and diversity, and MOEA/D is used to optimize the parameters of the model to enhance the generalization ability of the prediction model. Finally, the final load forecasting results are obtained by summing up the weighted outputs of each forecasting model with ensemble learning method. The experimental results show that compared with several current better load forecasting methods, this method has obvious advantages in prediction accuracy and generalization ability.
ISSN:0925-2312
1872-8286
DOI:10.1016/j.neucom.2020.01.031