Machine Learning Dynamic Ensemble Methods for Solar Irradiance and Wind Speed Predictions
This paper proposes to analyze the performance increase in the forecasting of solar irradiance and wind speed by implementing a dynamic ensemble architecture for intra-hour horizon ranging from 10 to 60 min for a 10 min time step data. Global horizontal irradiance (GHI) and wind speed were computed...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-10-01
|
| Series: | Atmosphere |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2073-4433/14/11/1635 |
| _version_ | 1797460170582786048 |
|---|---|
| author | Francisco Diego Vidal Bezerra Felipe Pinto Marinho Paulo Alexandre Costa Rocha Victor Oliveira Santos Jesse Van Griensven Thé Bahram Gharabaghi |
| author_facet | Francisco Diego Vidal Bezerra Felipe Pinto Marinho Paulo Alexandre Costa Rocha Victor Oliveira Santos Jesse Van Griensven Thé Bahram Gharabaghi |
| author_sort | Francisco Diego Vidal Bezerra |
| collection | DOAJ |
| description | This paper proposes to analyze the performance increase in the forecasting of solar irradiance and wind speed by implementing a dynamic ensemble architecture for intra-hour horizon ranging from 10 to 60 min for a 10 min time step data. Global horizontal irradiance (GHI) and wind speed were computed using four standalone forecasting models (random forest, k-nearest neighbors, support vector regression, and elastic net) to compare their performance against two dynamic ensemble methods, windowing and arbitrating. The standalone models and the dynamic ensemble methods were evaluated using the error metrics RMSE, MAE, R<sup>2</sup>, and MAPE. This work’s findings showcased that the windowing dynamic ensemble method was the best-performing architecture when compared to the other evaluated models. For both cases of wind speed and solar irradiance forecasting, the ensemble windowing model reached the best error values in terms of RMSE for all the assessed forecasting horizons. Using this approach, the wind speed forecasting gain was 0.56% when compared with the second-best forecasting model, whereas the gain for GHI prediction was 1.96%, considering the RMSE metric. The development of an ensemble model able to provide accurate and precise estimations can be implemented in real-time forecasting applications, helping the evaluation of wind and solar farm operation. |
| first_indexed | 2024-03-09T17:02:16Z |
| format | Article |
| id | doaj.art-2ae66d1f4b574d0e838d906a42a2d54d |
| institution | Directory Open Access Journal |
| issn | 2073-4433 |
| language | English |
| last_indexed | 2024-03-09T17:02:16Z |
| publishDate | 2023-10-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Atmosphere |
| spelling | doaj.art-2ae66d1f4b574d0e838d906a42a2d54d2023-11-24T14:28:30ZengMDPI AGAtmosphere2073-44332023-10-011411163510.3390/atmos14111635Machine Learning Dynamic Ensemble Methods for Solar Irradiance and Wind Speed PredictionsFrancisco Diego Vidal Bezerra0Felipe Pinto Marinho1Paulo Alexandre Costa Rocha2Victor Oliveira Santos3Jesse Van Griensven Thé4Bahram Gharabaghi5Department of Mechanical Engineering, Technology Center, Federal University of Ceará, Fortaleza 60020-181, CE, BrazilDepartment of Teleinformatics Engineering, Technology Center, Federal University of Ceará, Fortaleza 60020-181, CE, BrazilDepartment of Mechanical Engineering, Technology Center, Federal University of Ceará, Fortaleza 60020-181, CE, BrazilSchool of Engineering, University of Guelph, 50 Stone Rd. E, Guelph, ON N1G 2W1, CanadaSchool of Engineering, University of Guelph, 50 Stone Rd. E, Guelph, ON N1G 2W1, CanadaSchool of Engineering, University of Guelph, 50 Stone Rd. E, Guelph, ON N1G 2W1, CanadaThis paper proposes to analyze the performance increase in the forecasting of solar irradiance and wind speed by implementing a dynamic ensemble architecture for intra-hour horizon ranging from 10 to 60 min for a 10 min time step data. Global horizontal irradiance (GHI) and wind speed were computed using four standalone forecasting models (random forest, k-nearest neighbors, support vector regression, and elastic net) to compare their performance against two dynamic ensemble methods, windowing and arbitrating. The standalone models and the dynamic ensemble methods were evaluated using the error metrics RMSE, MAE, R<sup>2</sup>, and MAPE. This work’s findings showcased that the windowing dynamic ensemble method was the best-performing architecture when compared to the other evaluated models. For both cases of wind speed and solar irradiance forecasting, the ensemble windowing model reached the best error values in terms of RMSE for all the assessed forecasting horizons. Using this approach, the wind speed forecasting gain was 0.56% when compared with the second-best forecasting model, whereas the gain for GHI prediction was 1.96%, considering the RMSE metric. The development of an ensemble model able to provide accurate and precise estimations can be implemented in real-time forecasting applications, helping the evaluation of wind and solar farm operation.https://www.mdpi.com/2073-4433/14/11/1635wind energysolar energyrenewable energymachine learningforecasting ensembles |
| spellingShingle | Francisco Diego Vidal Bezerra Felipe Pinto Marinho Paulo Alexandre Costa Rocha Victor Oliveira Santos Jesse Van Griensven Thé Bahram Gharabaghi Machine Learning Dynamic Ensemble Methods for Solar Irradiance and Wind Speed Predictions Atmosphere wind energy solar energy renewable energy machine learning forecasting ensembles |
| title | Machine Learning Dynamic Ensemble Methods for Solar Irradiance and Wind Speed Predictions |
| title_full | Machine Learning Dynamic Ensemble Methods for Solar Irradiance and Wind Speed Predictions |
| title_fullStr | Machine Learning Dynamic Ensemble Methods for Solar Irradiance and Wind Speed Predictions |
| title_full_unstemmed | Machine Learning Dynamic Ensemble Methods for Solar Irradiance and Wind Speed Predictions |
| title_short | Machine Learning Dynamic Ensemble Methods for Solar Irradiance and Wind Speed Predictions |
| title_sort | machine learning dynamic ensemble methods for solar irradiance and wind speed predictions |
| topic | wind energy solar energy renewable energy machine learning forecasting ensembles |
| url | https://www.mdpi.com/2073-4433/14/11/1635 |
| work_keys_str_mv | AT franciscodiegovidalbezerra machinelearningdynamicensemblemethodsforsolarirradianceandwindspeedpredictions AT felipepintomarinho machinelearningdynamicensemblemethodsforsolarirradianceandwindspeedpredictions AT pauloalexandrecostarocha machinelearningdynamicensemblemethodsforsolarirradianceandwindspeedpredictions AT victoroliveirasantos machinelearningdynamicensemblemethodsforsolarirradianceandwindspeedpredictions AT jessevangriensventhe machinelearningdynamicensemblemethodsforsolarirradianceandwindspeedpredictions AT bahramgharabaghi machinelearningdynamicensemblemethodsforsolarirradianceandwindspeedpredictions |