Iterative multi-task learning for time-series modeling of solar panel PV outputs

Time-series modeling of PV output for solar panels can help solar panel owners understand the power systems’ time-varying behavior and be prepared for the load demand. The time-series forecast/prediction can become challenging due to many missing observations or a lack of historical records that are not sufficient to establish statistical models. Increasing PV measurement frequency over a longer period increases the cost in the detection of the PV fluctuation. This paper proposes an efficient approach to iterative multi-task learning for time series (MTL-GP-TS) that improves prediction of the PV output without increasing measurement efforts by sharing the information among PV data from multiple similar solar panels. The proposed iterative MTL-GP-TS model learns/imputes unobserved or missing values in a dataset of time series associated with the solar panel of interest to predict the PV trend. Additionally, the method improves and generalizes the traditional multi-task learning for Gaussian Process to the learning of both global trend and local irregular components in time series. A real-world case study demonstrated that the proposed method could result in substantial improvement of predictions over conventional approaches. The paper also discusses the selection of parameters and data sources when implementing the proposed algorithm.

[1]  Gordon Reikard.  Predicting solar radiation at high resolutions: A comparison of time series forecasts , 2009 .

[2]  Jian Su,et al.  A novel bidirectional mechanism based on time series model for wind power forecasting , 2016 .

[3]  Edwin V. Bonilla,et al.  Multi-task Gaussian Process Prediction , 2007, NIPS.

[4]  Ali Al-Alili,et al.  A new approach for model validation in solar radiation using wavelet, phase and frequency coherence analysis , 2016 .

[5]  Cyril Voyant,et al.  Forecasting of preprocessed daily solar radiation time series using neural networks , 2010 .

[6]  N. Altman.  An Introduction to Kernel and Nearest-Neighbor Nonparametric Regression , 1992 .

[7]  Yan Su,et al.  An ARMAX model for forecasting the power output of a grid connected photovoltaic system , 2014 .

[8]  Bruno Sinopoli,et al.  Kalman filtering with intermittent observations , 2004, IEEE Transactions on Automatic Control.

[9]  Zechun Hu,et al.  Photovoltaic and solar power forecasting for smart grid energy management , 2015 .

[10]  Samuel Kaski,et al.  Focused Multi-task Learning Using Gaussian Processes , 2011, ECML/PKDD.

[11]  L. D. Monache,et al.  An analog ensemble for short-term probabilistic solar power forecast , 2015 .

[12]  Stefan Pfenninger,et al.  Dealing with multiple decades of hourly wind and PV time series in energy models: A comparison of methods to reduce time resolution and the planning implications of inter-annual variability , 2017 .

[13]  Paras Mandal,et al.  Forecasting Power Output of Solar Photovoltaic System Using Wavelet Transform and Artificial Intelligence Techniques , 2012, Complex Adaptive Systems.

[14]  Guido Sanguinetti,et al.  Multi-task learning for pKa prediction , 2012, Journal of Computer-Aided Molecular Design.

[15]  Kasra Mohammadi,et al.  A support vector machine–firefly algorithm-based model for global solar radiation prediction , 2015 .

[16]  B. Abraham,et al.  A nonlinear time series model and estimation of missing observations , 1991 .

[17]  Guido Sanguinetti,et al.  Bayesian Multitask Classification With Gaussian Process Priors , 2011, IEEE Transactions on Neural Networks.

[18]  Le Xie,et al.  Multitime-Scale Data-Driven Spatio-Temporal Forecast of Photovoltaic Generation , 2015, IEEE Transactions on Sustainable Energy.

[19]  Hui Liu,et al.  Comparison of two new ARIMA-ANN and ARIMA-Kalman hybrid methods for wind speed prediction , 2012 .

[20]  Hui Liu,et al.  Forecasting models for wind speed using wavelet, wavelet packet, time series and Artificial Neural Networks , 2013 .

[21]  X. Wen,et al.  A wavelet-coupled support vector machine model for forecasting global incident solar radiation using limited meteorological dataset , 2016 .

[22]  O. Perpiñán,et al.  PV power forecast using a nonparametric PV model , 2015 .

[23]  Rich Caruana,et al.  Multitask Learning , 1997, Machine Learning.

[24]  Hisashi Kashima,et al.  Self-measuring Similarity for Multi-task Gaussian Process , 2011, ICML Unsupervised and Transfer Learning.

[25]  Jieping Ye,et al.  A transfer learning approach for network modeling , 2012, IIE transactions : industrial engineering research & development.

[26]  Cyril Voyant,et al.  Bayesian rules and stochastic models for high accuracy prediction of solar radiation , 2013, ArXiv.

[27]  H. Pedro,et al.  Assessment of forecasting techniques for solar power production with no exogenous inputs , 2012 .

[28]  Shai Ben-David,et al.  Exploiting Task Relatedness for Mulitple Task Learning , 2003, COLT.

[29]  G. King,et al.  What to Do about Missing Values in Time‐Series Cross‐Section Data , 2010 .