Genetic programming for photovoltaic plant output forecasting

Russo, Marco; Leotta, Giuseppe; Pugliatti, Paola Maria; Gigliucci, G.

doi:10.1016/j.solener.2014.02.021

Cited by 39 publications

(21 citation statements)

References 19 publications

(21 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…This section aims at describing briefly the support vector machine and firefly optimization algorithm as well as the encoding and methodology carried out to estimate the monthly mean daily global solar radiation with the proposed hybrid SVM-FFA approach. The description of ANN, GP, and ARMA can be found in the literature (Mora-López and Sidrach-de-Cardona 1998; Alam et al 2009;Şenkal and Kuleli 2009;Voyant et al 2012;Russo et al 2014). …”

Section: Methodsmentioning

confidence: 99%

RETRACTED ARTICLE: A hybrid SVM-FFA method for prediction of monthly mean global solar radiation

Shamshirband

Mohammadi

Chong

et al. 2015

Theor Appl Climatol

View full text Add to dashboard Cite

In this study, a hybrid support vector machine-firefly optimization algorithm (SVM-FFA) model is proposed to estimate monthly mean horizontal global solar radiation (HGSR). The merit of SVM-FFA is assessed statistically by comparing its performance with three previously used approaches. Using each approach and long-term measured HGSR, three models are calibrated by considering different sets of meteorological parameters measured for Bandar Abbass situated in Iran. It is found that the model (3) utilizing the combination of relative sunshine duration, difference between maximum and minimum temperatures, relative humidity, water vapor pressure, average temperature, and extraterrestrial solar radiation shows superior performance based upon all approaches. Moreover, the extraterrestrial radiation is introduced as a significant parameter to accurately estimate the global solar radiation. The survey results reveal that the developed SVM-FFA approach is greatly capable to provide favorable predictions with significantly higher precision than other examined techniques. For the SVM-FFA (3), the statistical indicators of mean absolute percentage error (MAPE), root mean square error (RMSE), relative root mean square error (RRMSE), and coefficient of determination (R 2 ) are 3.3252 %, 0.1859 kWh/m 2 , 3.7350 %, and 0.9737, respectively which according to the RRMSE has an excellent performance. As a more evaluation of SVM-FFA (3), the ratio of estimated to measured values is computed and found that 47 out of 48 months considered as testing data fall between 0.90 and 1.10. Also, by performing a further verification, it is concluded that SVM-FFA (3) offers absolute superiority over the empirical models using relatively similar input parameters. In a nutshell, the hybrid SVM-FFA approach would be considered highly efficient to estimate the HGSR.

show abstract

Section: Methodsmentioning

confidence: 99%

RETRACTED ARTICLE: A hybrid SVM-FFA method for prediction of monthly mean global solar radiation

Shamshirband

Mohammadi

Chong

et al. 2015

Theor Appl Climatol

View full text Add to dashboard Cite

show abstract

“…They conclude that weather data is likely to be helpful for forecasting horizons of several hours, while short term photovoltaic energy can be predicted only from previous energy data without weather information. However, some argue that clouds analysis from sky images is potentially useful for short term solar energy forecasting (Chow et al, 2011;Huang et al, 2014;Lipperheide et al, 2015;Russo et al, 2014;Urquhart et al, 2013). For this reason, sky images have been used in solar energy forecasting (Chow et al, 2011;Chu et al, 2015;Urquhart et al, 2013;Yang et al, 2014).…”

Section: Related Workmentioning

confidence: 99%

Deep photovoltaic nowcasting

et al. 2018

View full text Add to dashboard Cite

Predicting the short-term power output of a photovoltaic panel is an important task for the efficient management of smart grids. Short-term forecasting at the minute scale, also known as nowcasting, can benefit from sky images captured by regular cameras and installed close to the solar panel. However, estimating the weather conditions from these images-sun intensity, cloud appearance and movement, etc.-is a very challenging task that the community has yet to solve with traditional computer vision techniques. In this work, we propose to learn the relationship between sky appearance and the future photovoltaic power output using deep learning. We train several variants of convolutional neural networks which take historical photovoltaic power values and sky images as input and estimate photovoltaic power in a very short term future. In particular, we compare three different architectures based on: a multi-layer perceptron (MLP), a convolutional neural network (CNN), and a long short term memory (LSTM) module. We evaluate our approach quantitatively on a dataset of photovoltaic power values and corresponding images gathered in Kyoto, Japan. Our experiments reveal that the MLP network, already used similarly in previous work, achieves an RMSE skill score of 7% over the commonly-used persistence baseline on the 1-minute future photovoltaic power prediction task. Our CNN-based network improves upon this with a 12% skill score. In contrast, our LSTM-based model, which can learn the temporal dependencies in the data, achieves a 21% RMSE skill score, thus outperforming all other approaches.

show abstract

“…Other recent applications of GP in chemical engineering range from soft-sensor development for biochemical systems to prediction of the synthesis of heat-integrated complex distillation systems (Wang et al 2008). For more GP applications, see for example, Kovačič and Šarler (2014), Khadem et al (2014), Russo et al (2014) and Barmpalexis et al (2011).…”

Section: Genetic Programming (Gp)mentioning

confidence: 99%

Computational intelligence based models for prediction of elemental composition of solid biomass fuels from proximate analysis

Ghugare

Tiwary

Tambe

2014

Int J Syst Assur Eng Manag

View full text Add to dashboard Cite

Biomass is a renewable and sustainable source of ''green'' energy. The elemental composition comprising carbon (C), hydrogen (H) and oxygen (O) as major components, is an important measure of the biomass fuel's energy content. Its knowledge is also valuable in: (a) computing material balance in a biomass-based process, (b) designing and operating biomass utilizing efficient and clean combustors, gasifiers and boilers, (c) fixing the quantity of oxidants required for biomass combustion/ gasification, and (d) determining the volume and composition of the combustion/gasification gases. Obtaining the elemental composition of a biomass fuel via ultimate analysis is an expensive and time-consuming task. In comparison, proximate analysis that determines fixed carbon, ash, volatile matter and moisture content is a cruder characterization of the fuel and easier to perform. Thus, there exists a need for models possessing high accuracies for predicting the elemental composition of a solid biomass fuel from its proximate analysis constituents. Accordingly, this study utilizes three computational intelligence (CI) formalisms, namely, genetic programming, artificial neural networks and support vector regression, for developing nonlinear models for the prediction of C, H and O fractions of solid biomass fuels. A large database of 830 biomasses has been used in the stated model development. A comparison of the prediction accuracy and generalization performance of the nine CI-based models (three each for C, H and O) with that of the currently available linear models indicates that the CI-based models have consistently and significantly outperformed their linear counterparts. The models developed in this study have proved to be the best models for the prediction of elemental composition of solid biomass fuels from their proximate analyses.

show abstract

Genetic programming for photovoltaic plant output forecasting

Cited by 39 publications

References 19 publications

RETRACTED ARTICLE: A hybrid SVM-FFA method for prediction of monthly mean global solar radiation

RETRACTED ARTICLE: A hybrid SVM-FFA method for prediction of monthly mean global solar radiation

Deep photovoltaic nowcasting

Computational intelligence based models for prediction of elemental composition of solid biomass fuels from proximate analysis

Contact Info

Product

Resources

About