Probabilistic solar power forecasting using partially linear additive quantile regression models: an application to South African data
| dc.contributor.advisor | Sigauke, C. | |
| dc.contributor.advisor | Bere, A. | |
| dc.contributor.advisor | Mulaudzi, T, S, | |
| dc.contributor.author | Mpfumali, Phathutshedzo | |
| dc.date | 2019 | |
| dc.date.accessioned | 2019-06-05T13:16:19Z | |
| dc.date.available | 2019-06-05T13:16:19Z | |
| dc.date.issued | 2019-05-18 | |
| dc.description | MSc (Statistics) | en_US |
| dc.description | Department of Statistics | |
| dc.description.abstract | This study discusses an application of partially linear additive quantile regression models in predicting medium-term global solar irradiance using data from Tellerie radiometric station in South Africa for the period August 2009 to April 2010. Variables are selected using a least absolute shrinkage and selection operator (Lasso) via hierarchical interactions and the parameters of the developed models are estimated using the Barrodale and Roberts's algorithm. The best models are selected based on the Akaike information criterion (AIC), Bayesian information criterion (BIC), adjusted R squared (AdjR2) and generalised cross validation (GCV). The accuracy of the forecasts is evaluated using mean absolute error (MAE) and root mean square errors (RMSE). To improve the accuracy of forecasts, a convex forecast combination algorithm where the average loss su ered by the models is based on the pinball loss function is used. A second forecast combination method which is quantile regression averaging (QRA) is also used. The best set of forecasts is selected based on the prediction interval coverage probability (PICP), prediction interval normalised average width (PINAW) and prediction interval normalised average deviation (PINAD). The results show that QRA is the best model since it produces robust prediction intervals than other models. The percentage improvement is calculated and the results demonstrate that QRA model over GAM with interactions yields a small improvement whereas QRA over a convex forecast combination model yields a higher percentage improvement. A major contribution of this dissertation is the inclusion of a non-linear trend variable and the extension of forecast combination models to include the QRA. | en_US |
| dc.description.sponsorship | NRF | en_US |
| dc.format.extent | 1 online resource (xiv, 115 leaves) | |
| dc.identifier.apacitation | Mpfumali, P. (2019). <i>Probabilistic solar power forecasting using partially linear additive quantile regression models: an application to South African data</i>. (). . Retrieved from http://hdl.handle.net/11602/1349 | en_ZA |
| dc.identifier.chicagocitation | Mpfumali, Phathutshedzo. <i>"Probabilistic solar power forecasting using partially linear additive quantile regression models: an application to South African data."</i> ., , 2019. http://hdl.handle.net/11602/1349 | en_ZA |
| dc.identifier.citation | Mpfumali, Phathutshedzo (2019) Probabilistic solar power forecasting using partially linear additive quantile regression models: an application to South African data, University of Venda, South Africa,<http://hdl.handle.net/11602/1349>. | |
| dc.identifier.ris | TY - Dissertation AU - Mpfumali, Phathutshedzo AB - This study discusses an application of partially linear additive quantile regression models in predicting medium-term global solar irradiance using data from Tellerie radiometric station in South Africa for the period August 2009 to April 2010. Variables are selected using a least absolute shrinkage and selection operator (Lasso) via hierarchical interactions and the parameters of the developed models are estimated using the Barrodale and Roberts's algorithm. The best models are selected based on the Akaike information criterion (AIC), Bayesian information criterion (BIC), adjusted R squared (AdjR2) and generalised cross validation (GCV). The accuracy of the forecasts is evaluated using mean absolute error (MAE) and root mean square errors (RMSE). To improve the accuracy of forecasts, a convex forecast combination algorithm where the average loss su ered by the models is based on the pinball loss function is used. A second forecast combination method which is quantile regression averaging (QRA) is also used. The best set of forecasts is selected based on the prediction interval coverage probability (PICP), prediction interval normalised average width (PINAW) and prediction interval normalised average deviation (PINAD). The results show that QRA is the best model since it produces robust prediction intervals than other models. The percentage improvement is calculated and the results demonstrate that QRA model over GAM with interactions yields a small improvement whereas QRA over a convex forecast combination model yields a higher percentage improvement. A major contribution of this dissertation is the inclusion of a non-linear trend variable and the extension of forecast combination models to include the QRA. DA - 2019-05-18 DB - ResearchSpace DP - Univen KW - Lasso via hierarchical interaction KW - Partiallly linear additive KW - Models KW - Probabilistic solar power forecasting KW - Forecast combination LK - https://univendspace.univen.ac.za PY - 2019 T1 - Probabilistic solar power forecasting using partially linear additive quantile regression models: an application to South African data TI - Probabilistic solar power forecasting using partially linear additive quantile regression models: an application to South African data UR - http://hdl.handle.net/11602/1349 ER - | en_ZA |
| dc.identifier.uri | http://hdl.handle.net/11602/1349 | |
| dc.identifier.vancouvercitation | Mpfumali P. Probabilistic solar power forecasting using partially linear additive quantile regression models: an application to South African data. []. , 2019 [cited yyyy month dd]. Available from: http://hdl.handle.net/11602/1349 | en_ZA |
| dc.language.iso | en | en_US |
| dc.rights | University of Venda | |
| dc.subject | Lasso via hierarchical interaction | en_US |
| dc.subject | UCTD | en_ZA |
| dc.subject | Models | en_US |
| dc.subject | Probabilistic solar power forecasting | en_US |
| dc.subject | Forecast combination | en_US |
| dc.subject.ddc | 621.450968 | |
| dc.subject.lcsh | Solar energy -- South Africa | |
| dc.subject.lcsh | Solar energy -- Climatic factors -- South Africa | |
| dc.subject.lcsh | Solar power plants -- South Africa | |
| dc.subject.lcsh | Solar cells -- South Africa | |
| dc.title | Probabilistic solar power forecasting using partially linear additive quantile regression models: an application to South African data | en_US |
| dc.type | Dissertation | en_US |