Prediction Of Daily Global Solar Radiation Research Articles

Enhanced Daily Global Solar Radiation Prediction Through Hybrid Artificial Neural Network and Adaptive Neuro-Fuzzy Inference System with Meta-Heuristic Algorithm Integration

Enhanced Daily Global Solar Radiation Prediction Through Hybrid Artificial Neural Network and Adaptive Neuro-Fuzzy Inference System with Meta-Heuristic Algorithm Integration

A hybrid approach based on complete ensemble empirical mode decomposition with adaptive noise for multi-step-ahead solar radiation forecasting

Abstract Accurate measurements of solar radiation are required to ensure that power and energy systems continue to function effectively and securely. On the other hand, estimating it is extremely challenging due to the non-stationary behaviour and randomness of its components. In this research, a novel hybrid forecasting model, namely complete ensemble empirical mode decomposition with adaptive noise–Gaussian process regression (CEEMDAN–GPR), has been developed for daily global solar radiation prediction. The non-stationary global solar radiation series is transformed by CEEMDAN into regular subsets. After that, the GPR model uses these subsets as inputs to perform its prediction. According to the results of this research, the performance of the developed hybrid model is superior to two widely used hybrid models for solar radiation forecasting, namely wavelet–GPR and wavelet packet–GPR, in terms of mean square error, root mean square error, coefficient of determination and relative root mean square error values, which reached 3.23 MJ/m2/day, 1.80 MJ/m2/day, 95.56%, and 8.80%, respectively (for one-step forward forecasting). The proposed hybrid model can be used to ensure the safe and reliable operation of the electricity system.

Development of a hybrid computational intelligent model for daily global solar radiation prediction

Providing an accurate and reliable solar radiation prediction is highly significant for optimal design and management of thermal and solar photovoltaic systems. It is massively essential for producing renewable and clean energy production. In the current research, an exploration of newly developed computation intelligent models based on the hybridization of Covariance Matrix Adaptive Evolution Strategies (CMAES) with Extreme Gradient Boosting (XGB) and Multi-Adaptive Regression Splines (MARS) models for building robust predictive models for daily scale solar radiation. The proposed hybrid models are examined at four meteorological stations (i.e., Boromo, Dori, Gaoua, and Po) located in the Burkina Faso region, Sub-Sahara Africa. The prediction matrix is constructed using several related climate parameters. Based on the statistical and graphical evaluation, the proposed models demonstrated an outstanding predictability performance. The computational experiments showed that the proposed framework consistently enhanced prediction accuracy by 17.2%, 4.9%, 39.8%, and 44.5%, at Boromo, Dori, Gaoua, and Po stations using the Root Mean Squared Error over the testing phase against the existing methods proposed in the literature. The resulting method has the potential to be used as a surrogate tool for solar radiation prediction on the stations. The findings of this study contribute to a growing body of literature towards advancing the study of machine learning methods for predicting renewable energy production.

Model selection for accurate daily global solar radiation prediction in China

Solar energy can share with a big percentage in solving in the environmental pollution and energy crisis in China. For the most application of solar energy, an accurate information of solar radiation amount received by a horizontal surface is the first step. In this study, nine-day of the year based models are calibrated and evaluated to estimate the daily global solar radiation (DGSR) on a horizontal surface using long-term data of 84 stations all over China. After dividing China into five solar climatic zones, the highest performance model for each zone is chosen to be a general day of the year based models and then calibrate and evaluated at each solar zone. Based on statistical indicator; the root mean square error (RMSE), the mean absolute bias error (MABE), coefficient of determination (R2) and the correlation coefficient (r) and the uncertainty based on a 95% confidence level (U95) and Taylor diagram, results show that the models perform best in the solar zones that receive a high amount of solar radiation and the performance decreases with decreasing of the daily solar radiation incidence. The hybrid sine and cosine wave day of the year based models exhibit the best performance, model D7 is the best for zone I and model D8 for zones II and IV. For zones II and V, the Gaussian form model (D3 model) and the 4th order polynomial model (D4 model) yield high performance, respectively. Whereas, the two Gaussian form model is the worth performance for all zones. The developed general day of the year based models for each zone in this study can be used with high accuracy to estimate the (DGSR) on a horizontal surface in areas where there is no data or the data is missing all over China.

Potential of kernel-based nonlinear extension of Arps decline model and gradient boosting with categorical features support for predicting daily global solar radiation in humid regions

The knowledge of global solar radiation is of vital importance for the design and use of solar energy systems. This study evaluated the potential of two new powerful machine learning models, i.e., kernel-based nonlinear extension of Arps decline model and gradient boosting with categorical features support, for accurately estimating daily global solar radiation in humid regions. These two models were also compared with the multilayer perceptron, M5 model tree, random forest and multivariate adaptive regression spline models, using five input combinations of daily meteorological data during 2001–2015 from four weather stations in the (sub)tropical humid regions of South China. The results showed that, when lack of complete meteorological data, machine learning models using the ratio of actual and theoretical sunshine duration, maximum and minimum temperatures obtained satisfactory daily global solar radiation estimates. Generally, the kernel-based nonlinear extension of Arps decline model offered the best prediction accuracy among the studied models, followed by the gradient boosting with categorical features support. The multilayer perceptron model exhibited the smallest average percentage increase in the root mean square error during testing over the training values, followed by the kernel-based nonlinear extension of Arps decline model. Both the kernel-based nonlinear extension of Arps decline model and gradient boosting with categorical features support were successfully applied to develop general models for daily global solar radiation prediction (differences in root mean square error <5% compared with local models). The gradient boosting with categorical features support and multilayer perceptron exhibited much smaller computational time for both local and general models. Comprehensively considering the accuracy, stability and computational time, both the kernel-based nonlinear extension of Arps decline model and gradient boosting with categorical features support are recommended for predicting daily global solar radiation in the humid regions of China.

RETRACTED ARTICLE: Day of the year-based prediction of horizontal global solar radiation by a neural network auto-regressive model

The availability of accurate solar radiation data is essential for designing as well as simulating the solar energy systems. In this study, by employing the long-term daily measured solar data, a neural network auto-regressive model with exogenous inputs (NN-ARX) is applied to predict daily horizontal global solar radiation using day of the year as the sole input. The prime aim is to provide a convenient and precise way for rapid daily global solar radiation prediction, for the stations and their immediate surroundings with such an observation, without utilizing any meteorological-based inputs. To fulfill this, seven Iranian cities with different geographical locations and solar radiation characteristics are considered as case studies. The performance of NN-ARX is compared against the adaptive neuro-fuzzy inference system (ANFIS). The achieved results prove that day of the year-based prediction of daily global solar radiation by both NN-ARX and ANFIS models would be highly feasible owing to the accurate predictions attained. Nevertheless, the statistical analysis indicates the superiority of NN-ARX over ANFIS. In fact, the NN-ARX model represents high potential to follow the measured data favorably for all cities. For the considered cities, the attained statistical indicators of mean absolute bias error, root mean square error, and coefficient of determination for the NN-ARX models are in the ranges of 0.44–0.61 kWh/m2, 0.50–0.71 kWh/m2, and 0.78–0.91, respectively.

Daily global solar radiation prediction based on a hybrid Coral Reefs Optimization – Extreme Learning Machine approach

This paper discusses the performance of a novel Coral Reefs Optimization – Extreme Learning Machine (CRO–ELM) algorithm in a real problem of global solar radiation prediction. The work considers different meteorological data from the radiometric station at Murcia (southern Spain), both from measurements, radiosondes and meteorological models, and fully describes the hybrid CRO–ELM to solve the prediction of the daily global solar radiation from these data. The algorithm is designed in such a way that the ELM solves the prediction problem, whereas the CRO evolves the weights of the neural network, in order to improve the solutions obtained. The experiments carried out have shown that the CRO–ELM approach is able to obtain an accurate prediction of the daily global radiation, better than the classical ELM, and the Support Vector Regression algorithm.

The potential of different artificial neural network (ANN) techniques in daily global solar radiation modeling based on meteorological data

The main objective of present study is to predict daily global solar radiation (GSR) on a horizontal surface, based on meteorological variables, using different artificial neural network (ANN) techniques. Daily mean air temperature, relative humidity, sunshine hours, evaporation, and wind speed values between 2002 and 2006 for Dezful city in Iran (32°16′N, 48°25′E), are used in this study. In order to consider the effect of each meteorological variable on daily GSR prediction, six following combinations of input variables are considered: (I) Day of the year, daily mean air temperature and relative humidity as inputs and daily GSR as output. (II) Day of the year, daily mean air temperature and sunshine hours as inputs and daily GSR as output. (III) Day of the year, daily mean air temperature, relative humidity and sunshine hours as inputs and daily GSR as output. (IV) Day of the year, daily mean air temperature, relative humidity, sunshine hours and evaporation as inputs and daily GSR as output. (V) Day of the year, daily mean air temperature, relative humidity, sunshine hours and wind speed as inputs and daily GSR as output. (VI) Day of the year, daily mean air temperature, relative humidity, sunshine hours, evaporation and wind speed as inputs and daily GSR as output. Multi-layer perceptron (MLP) and radial basis function (RBF) neural networks are applied for daily GSR modeling based on six proposed combinations. The measured data between 2002 and 2005 are used to train the neural networks while the data for 214 days from 2006 are used as testing data. The comparison of obtained results from ANNs and different conventional GSR prediction (CGSRP) models shows very good improvements (i.e. the predicted values of best ANN model (MLP-V) has a mean absolute percentage error (MAPE) about 5.21% versus 10.02% for best CGSRP model (CGSRP 5)).