Analysis of daily solar power prediction with data-driven approaches

• Published in: Applied energy Vol. 126; pp. 29 - 37
• Main Authors: Long, Huan, Zhang, Zijun, Su, Yan
• Format: Journal Article
• Language: English
• Published: Kidlington Elsevier Ltd 01.08.2014; Elsevier
• Subjects: Applied sciences; Artificial Neural Network (ANN); Data mining; Energy; Exact sciences and technology; meteorological parameters; neural networks; prediction; regression analysis; solar energy; Solar power prediction; Support Vector Machine (SVM); support vector machines; Time-series model; Solar power prediction; Data mining; Time-series model; Support Vector Machine (SVM); Artificial Neural Network (ANN); Solar energy; Time series; Models; Electric power production; Neural network; Modeling
• ISSN: 0306-2619; 1872-9118
• DOI: 10.1016/j.apenergy.2014.03.084

•We develop daily solar power prediction models with data-driven approaches.•We introduce a parameter selection procedure for reducing dimensions of prediction models.•We report a comparative analysis of data mining algorithms in daily solar power prediction.•Data mining algorithms can perform better than persistent methods.•None of data mining algorithms can dominate others in all prediction scenarios. Daily solar power prediction using data-driven approaches is studied. Four famous data-driven approaches, the Artificial Neural Network (ANN), the Support Vector Machine (SVM), the k-nearest neighbor (kNN), and the multivariate linear regression (MLR), are applied to develop the prediction models. The persistent model is considered as a baseline for evaluating the effectiveness of data-driven approaches. A procedure of selecting input parameters for solar power prediction models is addressed. Two modeling scenarios, including and excluding meteorological parameters as inputs, are assessed in the model development. A comparative analysis of the data-driven algorithms is conducted. The capability of data-driven models in multi-step ahead prediction is examined. The computational results indicate that none of the algorithms can outperform others in all considered prediction scenarios.