A deep learning-based evolutionary model for short-term wind speed forecasting: a case study of the Lillgrund offshore wind farm

Mehdi Neshat, Meysam Majidi Nezhad, Ehsan Abbasnejad, Seyedali Mirjalili, Lina Bertling Tjernberg, Davide Astiaso Garcia, Bradley Alexander, Markus Wagner

Research output: Contribution to journalArticleResearchpeer-review

130 Citations (Scopus)


Due to expanding global environmental issues and growing energy demand, wind power technologies have been studied extensively. Accurate and robust short-term wind speed forecasting is crucial for large-scale integration of wind power generation into the power grid. However, the seasonal and stochastic characteristics of wind speed make forecasting a challenging task. This study adopts a novel hybrid deep learning-based evolutionary approach in an attempt to improve the accuracy of wind speed prediction. This hybrid model consists of a bidirectional long short-term memory neural network, an effective hierarchical evolutionary decomposition technique and an improved generalised normal distribution optimisation algorithm for hyper-parameter tuning. The proposed hybrid approach was trained and tested on data gathered from an offshore wind turbine installed in a Swedish wind farm located in the Baltic Sea with two forecasting horizons: ten-minutes ahead and one-hour ahead. The experimental results indicated that the new approach is superior to six other applied machine learning models and a further seven hybrid models, as measured by seven performance criteria.

Original languageEnglish
Article number114002
Number of pages25
JournalEnergy Conversion and Management
Publication statusPublished - 15 May 2021
Externally publishedYes


  • Deep learning models
  • Evolutionary algorithms
  • Generalised normal distribution optimisation
  • Hybrid evolutionary deep learning method
  • Short-term forecasting
  • Wind speed prediction

Cite this