Assessing ANN Architectures for Wind Turbine Power Prediction: A Comparative Study

Fethi Achouri, Mehdi Damou, Fouzi Harrou, Ying Sun, Benamar Bouyeddou

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

1 Scopus citations

Abstract

With the increasing importance of renewable energy sources, wind power has emerged as a significant contributor to the global energy mix. Accurate prediction of wind power production is essential for efficient grid integration and optimal utilization of wind resources. This study presents an investigation into the performance of various artificial neural network (ANN) models for wind power prediction. Different ANN architectures, including narrow, medium, and wide networks, as well as bilayered and trilayered structures, were explored to understand their impact on predictive capabilities. The data used for evaluation was collected from a 2.05 MW Senvion MM82 wind turbine. The performance of the ANN models was compared with Linear Regression (LR), Interactions LR, Robust LR, and Stepwise LR methods. Results revealed that the bilayered and trilayered ANN s achieved the best performance in wind power prediction. This study highlights the potential of ANN models in accurately predicting wind power, thereby facilitating efficient and reliable wind farm operations.

Original languageEnglish (US)
Title of host publication2023 International Conference on Decision Aid Sciences and Applications, DASA 2023
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages284-289
Number of pages6
ISBN (Electronic)9798350342055
DOIs
StatePublished - 2023
Event2023 International Conference on Decision Aid Sciences and Applications, DASA 2023 - Annaba, Algeria
Duration: Sep 16 2023Sep 17 2023

Publication series

Name2023 International Conference on Decision Aid Sciences and Applications, DASA 2023

Conference

Conference2023 International Conference on Decision Aid Sciences and Applications, DASA 2023
Country/TerritoryAlgeria
CityAnnaba
Period09/16/2309/17/23

Keywords

  • artificial neural networks
  • power prediction
  • regression learning
  • Wind turbines

ASJC Scopus subject areas

  • Computer Science Applications
  • Decision Sciences (miscellaneous)
  • Information Systems and Management
  • Computational Mathematics
  • Control and Optimization
  • Health Informatics

Fingerprint

Dive into the research topics of 'Assessing ANN Architectures for Wind Turbine Power Prediction: A Comparative Study'. Together they form a unique fingerprint.

Cite this