Testing machine learning approaches for wind plants power output

Alexander Malakhov, Fyodor Goncharov, Elena Gryazina

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

1 Citation (Scopus)

Abstract

Wind plant power output behaves like a function that strongly depends on the value and direction of wind speed, but the weather conditions have plenty of parameters that can affect the output. In this work, we are aimed to merge two different datasets to get the application for a wind turbine model. The first dataset contains weather parameters while the second consists of the wind power production for a region. Then we estimate the parameters of an equivalent wind turbine to evaluate physical model baseline. On the next step, we build and tune regression machine learning models: the gradient boosting with decision trees, Gaussian process, neural network and support vector machine. Additionally, we perform the sensitivity analysis for the most accurate trained models and get the most reliable weather parameters.

Original languageEnglish
Title of host publicationProceedings of the 1st IEEE 2019 International Youth Conference on Radio Electronics, Electrical and Power Engineering, REEPE 2019
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781538693346
DOIs
Publication statusPublished - 6 May 2019
Event1st IEEE International Youth Conference on Radio Electronics, Electrical and Power Engineering, REEPE 2019 - Moscow, Russian Federation
Duration: 14 Mar 201915 Mar 2019

Publication series

NameProceedings of the 1st IEEE 2019 International Youth Conference on Radio Electronics, Electrical and Power Engineering, REEPE 2019

Conference

Conference1st IEEE International Youth Conference on Radio Electronics, Electrical and Power Engineering, REEPE 2019
Country/TerritoryRussian Federation
CityMoscow
Period14/03/1915/03/19

Keywords

  • machine learning
  • power systems
  • RES
  • Wind energy
  • wind turbine

Fingerprint

Dive into the research topics of 'Testing machine learning approaches for wind plants power output'. Together they form a unique fingerprint.

Cite this