Novel Comparison of Machine Learning Techniques for Predicting Photovoltaic Output Power

Souhaila CHAHBOUN, Mohamed MAAROUFI

Abstract


Connected objects, Big data and Artificial intelligence are the main drivers of the global industrial revolutions. Among these digital technologies, Artificial intelligence has become a must in recent years. If the shift towards artificial intelligence-based solutions is only happening now, it is thanks to the explosion of computing power of computers, the introduction of new algorithms enabling to take advantage of Machine Learning's analysis and prediction capabilities along with the volume of available data generated in real time by sensors and Internet of Things activators. Artificial intelligence is already revolutionizing several industrial sectors. In the energy field, its ability to learn from data allows for the optimization of both energy generation and consumption. Energy needs are better assessed in order to enhance the energy transition. Therefore, this study aims to explore the development potential of artificial intelligence solutions in the energy sector. A performance comparison of several up-to-date Machine learning algorithms is conducted for the hourly prediction of the resulting power of photovoltaic panels. The model offering the best accuracy is identified using the most common performance metrics including R-squared, Root Mean Square Error, and Mean Absolute Error. Moreover, this paper investigates the prevailing inputs affecting the PV power through using correlation analysis. The methods investigated are Partial Least squares regression, LASSO, Multivariate Adaptive Regression Splines, Support vector regression, K-nearest Neighbor, Gradient Boosting, Random Forest and Bayesian regularized neural networks. Finally, residual analysis and diagnostic techniques are carried out to visually examine the fit of regression models.


Keywords


Artificial intelligence; Machine learning; Forecasting; Solar power; Photovoltaic output power

Full Text:

PDF

References


U. K. Das et al., “Forecasting of photovoltaic power generation and model optimization: A review,” Renew. Sustain. Energy Rev., vol. 81, no. April 2017, pp. 912–928, 2018, doi: 10.1016/j.rser.2017.08.017.

J. P. Lai, Y. M. Chang, C. H. Chen, and P. F. Pai, “A survey of machine learning models in renewable energy predictions,” Appl. Sci., vol. 10, no. 17, 2020, doi: 10.3390/app10175975.

S. Zhang, J. Wang, H. Liu, J. Tong, and Z. Sun, “Prediction of energy photovoltaic power generation based on artificial intelligence algorithm,” Neural Comput. Appl., vol. 0123456789, 2020, doi: 10.1007/s00521-020-05249-z.

S. Moslehi, T. A. Reddy, and S. Katipamula, “Evaluation of data-driven models for predicting solar photovoltaics power output,” Energy, vol. 142, pp. 1057–1065, 2018, doi: 10.1016/j.energy.2017.09.042.

S. Kanwal, B. Khan, S. M. Ali, C. A. Mehmood, and M. Q. Rauf, “Support vector machine and Gaussian process regression based modeling for photovoltaic power prediction,” Proc. - 2018 Int. Conf. Front. Inf. Technol. FIT 2018, pp. 117–122, 2019, doi: 10.1109/FIT.2018.00028.

Y. Li, Y. He, Y. Su, and L. Shu, “Forecasting the daily power output of a grid-connected photovoltaic system based on multivariate adaptive regression splines,” Appl. Energy, vol. 180, pp. 392–401, 2016, doi: 10.1016/j.apenergy.2016.07.052.

L. Gigoni et al., “Day-Ahead Hourly Forecasting of Power Generation from Photovoltaic Plants,” IEEE Trans. Sustain. Energy, vol. 9, no. 2, pp. 831–842, 2018, doi: 10.1109/TSTE.2017.2762435.

M. Kayri, I. Kayri, and M. T. Gencoglu, “The performance comparison of Multiple Linear Regression, Random Forest and Artificial Neural Network by using photovoltaic and atmospheric data,” 2017 14th Int. Conf. Eng. Mod. Electr. Syst. EMES 2017, pp. 1–4, 2017, doi: 10.1109/EMES.2017.7980368.

N. A. Ramli, M. F. A. Hamid, N. H. Azhan, and M. A. A. S. Ishak, “Solar power generation prediction by using k-nearest neighbor method,” AIP Conf. Proc., vol. 2129, no. July, 2019, doi: 10.1063/1.5118124.

A. Golder, J. Jneid, J. Zhao, and F. Bouffard, “Machine learning-based demand and PV power forecasts,” 2019 IEEE Electr. Power Energy Conf. EPEC 2019, vol. 3, 2019, doi: 10.1109/EPEC47565.2019.9074819.

N. Maitanova et al., “A machine learning approach to low-cost photovoltaic power prediction based on publicly available weather reports,” Energies, vol. 13, no. 3, 2020, doi: 10.3390/en13030735.

R. CoreTeam, R: A Language and Environment for Statistical Computing, vol. 2. 2018.

J. ying Zhang, Z. Wang, and X. han Zhang, “Research on photovoltaic output power short term prediction method based on machine learning,” Energy Syst., no. 0123456789, 2020, doi: 10.1007/s12667-020-00386-9.

L. Visser, T. Alskaif, and W. Van Sark, “Benchmark analysis of day-ahead solar power forecasting techniques using weather predictions,” Conf. Rec. IEEE Photovolt. Spec. Conf., pp. 2111–2116, 2019, doi: 10.1109/PVSC40753.2019.8980899.

E. Mower and R. Tibshirani, “Machine Learning and Big Data Analytics Section 5 Lasso and Ridge Regression,” no. October, pp. 1–2, 2018.

L. Massidda and M. Marrocu, “Use of Multilinear Adaptive Regression Splines and numerical weather prediction to forecast the power output of a PV plant in Borkum, Germany,” Sol. Energy, vol. 146, pp. 141–149, 2017, doi: 10.1016/j.solener.2017.02.007.

M. H. D. M. Ribeiro and L. dos Santos Coelho, “Ensemble approach based on bagging, boosting and stacking for short-term prediction in agribusiness time series,” Appl. Soft Comput. J., vol. 86, p. 105837, 2020, doi: 10.1016/j.asoc.2019.105837.

M. Malvoni and Y. Chaibi, “Machine learning based approaches for modeling the output power of photovoltaic array in real outdoor conditions,” Electron., vol. 9, no. 2, 2020, doi: 10.3390/electronics9020315.

C. Voyant et al., “Machine learning methods for solar radiation forecasting: A review,” Renew. Energy, vol. 105, pp. 569–582, 2017, doi: 10.1016/j.renene.2016.12.095.

J. M. M. Arce and E. Q. B. Macabebe, “Real-Time Power Consumption Monitoring and Forecasting Using Regression Techniques and Machine Learning Algorithms,” Proc. - 2019 IEEE Int. Conf. Internet Things Intell. Syst. IoTaIS 2019, pp. 135–140, 2019, doi: 10.1109/IoTaIS47347.2019.8980380.

A. El Bakri, Y. Berrada, and I. Boumhidi, “Bayesian regularized artificial neural network for fault detection and isolation in wind turbine,” 2017 Intell. Syst. Comput. Vision, ISCV 2017, 2017, doi: 10.1109/ISACV.2017.8054977.

L. Massidda and M. Marrocu, “Use of Multilinear Adaptive Regression Splines and numerical weather prediction to forecast the power output of a PV plant in Borkum, Germany,” Sol. Energy, vol. 146, pp. 141–149, 2017, doi: 10.1016/j.solener.2017.02.007.

A. Elamim, B. Hartiti, A. Barhdadi, A. Haibaoui, A. Lfakir, and P. Thevenin, “Photovoltaic output power forecast using artificial neural networks,” J. Theor. Appl. Inf. Technol., vol. 96, no. 15, pp. 5116–5126, 2018.

A. A. Babatunde and S. Abbasoglu, “Predictive analysis of photovoltaic plants specific yield with the implementation of multiple linear regression tool,” Environ. Prog. Sustain. Energy, vol. 38, no. 4, pp. 1–9, 2019, doi: 10.1002/ep.13098.

B. Greenwell, B. Boehmke, and J. Cunningham, “Package ‘gbm’ - Generalized Boosted Regression Models,” CRAN Repos., p. 39, 2020, [Online]. Available: https://cran.r-project.org/web/packages/gbm/gbm.pdf%0Ahttps://github.com/gbm-developers/gbm.

L. Breiman, A. Cutler, A. Liaw, and M. Wiener, “Package ‘randomForest’ - Breiman and Cutler’s Random Forests for Classification and Regression,” CRAN Repos., p. 29, 2018, doi: 10.1023/A:1010933404324.

P. Rodriguez, D. Gianola, and M. Paulino, “Package ‘brnn’ Title Bayesian Regularization for Feed-Forward Neural Networks,” p. 23, 2020.

A. Gosiewska and P. Biecek, “Auditor: An R package for model-agnostic visual validation and diagnostics,” R J., vol. 11, no. 2, pp. 85–98, 2020, doi: 10.32614/rj-2019-036.

J. Hernández-Orallo, “ROC curves for regression,” Pattern Recognit., vol. 46, no. 12, pp. 3395–3411, 2013, doi: 10.1016/j.patcog.2013.06.014.

J. Bi and K. P.Bennett, “Regression error characteristic curves,” Twent. Int. Conf. Mach. Learn. (ICML-2003), Washington, DC, 2003.

M. Kayri, “Predictive abilities of Bayesian regularization and levenberg-marquardt algorithms in artificial neural networks: A comparative empirical study on social data,” Math. Comput. Appl., vol. 21, no. 2, 2016, doi: 10.3390/mca21020020.




DOI (PDF): https://doi.org/10.20508/ijrer.v11i3.12056.g8252

Refbacks

  • There are currently no refbacks.


Online ISSN: 1309-0127

Publisher: Gazi University

IJRER is cited in SCOPUS, EBSCO, WEB of SCIENCE (Clarivate Analytics);

IJRER has been cited in Emerging Sources Citation Index from 2016 in web of science.

WEB of SCIENCE in 2025; 

h=35,

Average citation per item=6.59

Last three Years Impact Factor=(1947+1753+1586)/(146+201+78)=5286/425=12.43

Category Quartile:Q4