Solar Radiation Prediction Models Research Articles

Temporal Evolution, Oscillation and Coherence Characteristics Analysis of Global Solar Radiation Distribution in Major Cities in China’s Solar-Energy-Available Region Based on Continuous Wavelet Transform

The majority of the energy required for human survival is derived either directly or indirectly from solar radiation, thus it is important to investigate the periodic fluctuations in global solar radiation over time. In this study, six cities—Harbin, Shenyang, Beijing, Shanghai, Wuhan, and Guangzhou—located in the utilizable areas of solar energy in China, were selected, and the periodicity of the daily global solar radiation of these six cities with time were investigated by means of wavelet power spectrum analysis. Furthermore, Harbin, which has the lowest monthly average of global solar radiation in the cold of winter, was selected to explore the temporal evolution relationship between global solar radiation and other meteorological factors, and air quality factors by wavelet coherence analysis. Among the meteorological factors, the correlation between global solar radiation and sunshine duration exhibits the highest level of consistency. Global solar radiation demonstrates a positive association with atmospheric temperature/wind speed/precipitation over the annual cycle. Conversely, it displays a negative correlation with atmospheric pressure during this time frame. Additionally, on shorter timescales, global solar radiation is negatively correlated with air humidity and precipitation. Among air quality factors, the relationship between global solar radiation and CO/NO2/O3/PM2.5/PM10/SO2 primarily manifests in the broader annual cycle time scale. Within this context, global solar radiation shows a positive correlation with O3, while displaying negative associations with the other five air quality indicators. These findings contribute to urban environmental planning and the effective management and utilization of solar radiation, thereby providing valuable insights to guide the future development of cross-regional comprehensive energy utilization projects under diverse climatic and geographical conditions. Additionally, they serve as a reference for subsequent research aimed at enhancing the accuracy of global solar radiation prediction models.

An integrated skip convolutional network with residual learning and feature extraction for point and interval prediction of solar radiation

Spinning reserve based on solar radiation prediction can ensure the secure operation of large-scale grid-connected photovoltaic systems, but irrational spinning reserve can lead to substantial economic losses and even grid collapse. Therefore, it is imperative to identify characteristic patterns of solar radiation and establish an effective solar radiation prediction model. However, existing hybrid models often struggle to efficiently recognize and leverage input features, compromising the robustness of the model. To address this limitation, this study proposes an integrated skip-convolutional network with residual learning and feature extraction (InSCNet), which enhances the capability to represent information data and thereby improves the stability and accuracy of solar radiation prediction by employing a series of feature extraction and prediction blocks with residual learning. InSCNet includes a sophisticated feature extraction module to effectively address the underutilization of feature information, and it incorporates residual learning, skip-convolution, and recursive networks in the prediction module, reducing the risk of gradient vanishing or gradient explosion while enhancing prediction accuracy. Additionally, an interval estimation method with adaptively chosen distributions is introduced, which extends the prediction interval estimation method. The proposed InSCNet is rigorously evaluated using datasets from three major cities in Pakistan. Experimental results demonstrate that InSCNet outperforms existing solutions in both point and interval predictions.

Using of the Weibull distribution in developing global solar radiation forecasting models

AbstractIn this study, a solar radiation prediction model was developed using the Weibull distribution function (WDF). The main purpose of this study is to develop a new model for solar radiation (SR) estimation using the WDF and to bring this model to the literature. Although the WDF is widely used in wind energy forecasting due to its compatibility with wind speed data, it has not been used before in solar radiation forecasting model development. In this study, it is aimed to make SR estimation with the WDF for the first time. SR data for all regions where this new model was developed and its performance was examined were obtained from the General Directorate of Meteorology. In order to decide the success of these developed models, five different statistical metrics (RPE, MPE, MAPE, SSRE, and t‐stat) were discussed in the study. When the results are evaluated in general, it is seen that the new developed model has an acceptable performance. According to all test results in the region where the model was developed, it is seen that the new developed model showed the best performance with 0.0353, 0.4068, 7.1851, 0.1144, and 0.0162 values, respectively. The performance of this new developed model was observed in three different regions. When the statistical test results in these regions were examined, it was seen that the new model developed had a similar performance to the popular solar forecasting models widely used in the literature.

Solar Radiation Forecasting Models and their Thermodynamic Analysis in Asaba: Least Square Regression and Machine Learning Approach

Least square regression and machine learning tools were used for the development of global solar radiation forecasting models for Asaba region. Data from the year 2013-2022 from Nigerian Meteorological Agency, Asaba was used for this study. The least square regression method was used to develop four global solar radiation -based models, tagged H1, H2, H3 and H4 with characteristic day length, solar declination angle, rainfall amount, etc. as its model terms while the machine learning models produced multilayer perceptron, coarse Gaussian model (SVM-based model) and XGBoost model. The prediction factors like mean bias error, mean percentage error, root mean square error, Nash-Sutcliffe equation, coefficient of correlation (R), t-test, and coefficient of determination (R2) were considered using the model terms. The results indicates that H4 model outperformed H1, H2, H3, machine learning models (SVM-based model, multilayer perceptron and XGBoost) and other existing models (MA-MME and MLR) with a mean percentage error value of 0.740, RMSE value of 46.588, Nash-Sutcliffe equation value of 0.739, higher R2 value of 0.7391, t-test value of 2.595E-24 and mean bias error value of -6.88E-12. Thus, H4 model results fell within accepted range. Additionally, the exergy of the global solar radiation of Asaba varied from 20-185 W/m2 which are good. This shows that a more efficient and ideal global solar radiation prediction model (H4) has been developed for Asaba and other regions that share similar climatic conditions.

Multi-step solar radiation prediction using transformer: A case study from solar radiation data in Tokyo

The widespread advancement of computer technology resulted in the increasing usage of deep learning models for predicting solar radiation. Numerous studies have been conducted to explore their research potential. Nevertheless, the application of deep learning models in optimizing building energy systems, particularly in a multi-step solar radiation prediction model for model predictive control (MPC), remains a challenging task. This is mainly due to the intricacy of the time series and the possibility of accumulating errors in multistep forecasts. In this study, we propose the development of a transformer-based attention model for predicting multi-step solar irradiation at least 24 h in advance. The model is trained and tested using measured solar irradiation data and temperature forecast data obtained from the Tokyo Meteorological Agency. The findings indicate that the transformer model has the capability to effectively mitigate the issue of error accumulation. Additionally, the generative model exhibits a significant improvement in accuracy, with a 62.35% increase when compared to the conventional regression LSTM model. Additionally, the transformer model has been shown to attain superior prediction stability, mitigate the effects of error accumulation in multi-step forecasting, and circumvent training challenges stemming from gradient propagation issues that can occur with recurrent neural networks.

Photovoltaic capacity dynamic tracking model predictive control strategy of air-conditioning systems with consideration of flexible loads

Building photovoltaic (PV) power generation is intermittent, volatile, random, and uncontrollable; thus, the use of solar grid-connected power generation can lead to a series of problems, such as grid voltage fluctuations and power imbalance. As a typical flexible load, the scheduling and regulation of the heating, ventilation, and air conditioning (HVAC) system load can help a grid-connected solar grid achieve balanced and flexible operation. This study proposes a PV capacity dynamic tracking model predictive control strategy for air-conditioning systems with flexible loads. This strategy aims to effectively address the issue of unstable grid-connected output of solar PV systems. In particular, a solar radiation prediction model based on a long short-term memory neural network optimized by the grey wolf optimization algorithm was established, which effectively improved the model prediction accuracy. The cost function of HVAC flexible load control considers the law of solar PV output and the human thermal comfort model. A genetic algorithm (GA) is used to obtain the optimal control parameters for dynamic regulation. The GA was employed to efficiently and effectively optimize the control of flexible HVAC loads, considering the variability of the solar PV output and ensuring human thermal comfort. A dynamic regulation model was developed for a typical office building with a grid-connected solar PV power generation system. TRNSYS was utilized for verification based on the actual measured data and relevant meteorological parameters of the case building. A new evaluation metric, volatility, was proposed to evaluate net load fluctuation. The simulation platform based on the physical building yield results indicated a strong resemblance between the energy consumption curve under the predicted control conditions using the flexible model and the PV generation curve. Additionally, the net load volatility was measured to be 2.63. Compared with the predicted control condition without the flexible model, the net load volatility of the grid was reduced by 47.08%, and the energy-saving rate reached 10.89% in the summer.

An hourly solar radiation prediction model using eXtreme gradient boosting algorithm with the effect of fog-haze

An hourly solar radiation prediction model using eXtreme gradient boosting algorithm with the effect of fog-haze

Integration of satellite imagery and meteorological data to estimate solar radiation using machine learning models

Knowing the behavior of solar energy is imperative for its use in photovoltaic systems; moreover, the number of weather stations is insufficient. This study presents a method for the integration of solar resource data: images and datasets.  For this purpose, variables are extracted from images obtained from the GOES-13 satellite and integrated with variables obtained from meteorological stations. Subsequently, this data integration was used to train solar radiation prediction models in three different scenarios with data from 2012 and 2017. The predictive ability of five regression methods was evaluated, of which, neural networks had the highest performance in the scenario that integrates the meteorological variables and features obtained from the images. The analysis was performed using four evaluation metrics in each year. In the 2012 dataset, an R2 of 0.88 and an RMSE of 90.99 were obtained. On the other hand, in the 2017 dataset, an R2 of 0.92 and an RMSE of 40.97 were achieved. The model integrating data improves performance by up to 4% in R2 and up to 10 points less in the level of dispersion according to RMSE, with respect to models using separate data.

Multi-model approach applied to meteorological data for solar radiation forecasting using data-driven approaches

The accuracy of the solar forecasting model depends upon various factors such as attributes of the data, time horizon, time interval, length of the data, etc. This means attributes of data selection are an important feature of any solar forecasting model. A lot of research has been done on solar forecasting using deep learning and machine learning approaches, providing better accuracy for a particular type of model and data. However, how to accurately apply the multi-model techniques in the data-driven approaches of the solar radiation prediction model is still a challenging issue due to the complexity of the temporal resolution, length, and attribute selection of the data. This paper proposes a multi-model technique using data-driven approaches to get an improvised solar forecasting model with respect to the attributes of data. In this research, a multi-model method based on data-driven approaches is developed for short-term solar radiation forecasting. Three years of weather data with different temporal resolutions taken from reliable online sources were used for training and testing in this work. The weather data considered various input parameters such as temperature, humidity, pressure, wind speed, and wind direction; solar radiation data is considered the model’s target variable. This work is a two-stage process. In the first stage, multi-model methods are analyzed based on solar forecasts at different temporal resolutions, and best-performing models are selected. The optimization technique is applied to the best model in the second stage to obtain an improvised solar forecasting model. It is observed that the multi-model forecasts provide better accuracy in terms of R2, nRMSE, and MAE than the conventional forecasting model. In this work, the performance parameters of nRMSE = 0.34, MAE = 4.71, and R2 = 0.93 are obtained before optimization, and nRMSE = 0.25, MAE = 1.39 and R2 = 0.96 are obtained after optimization in the testing stage for an hourly dataset. For daily datasets, the performance parameters nRMSE = 0.24, MAE = 3.19, and R2 = 0.93 are obtained before optimization, and nRMSE = 0.16, MAE = 2.61, and R2 = 0.95 are obtained after optimization in the testing stage. As a result, the proposed algorithm focused on an efficient model to predict solar radiation. Accurate solar radiation prediction models are highly in demand in various smart grid applications. Therefore, the proposed work may help in suggesting a suitable model for solar energy forecasting.

Arithmetic optimization with hybrid deep learning algorithm based solar radiation prediction model

Solar radiation affects extreme weather occurrences, as well as on the temperature and mean sea level. Therefore, precise studies and measurements of geographical and temporal variations of solar radiation are required. The development of deep learning and machine learning methods for developing solar radiation predictive models is gaining traction. Therefore, this paper presents an Arithmetic Optimization with Hybrid Deep Learning based Solar Radiation Prediction model. The presented AOHDL-SRP model follows a three-stage process: pre-processing, prediction, and hyperparameter optimization. Primarily, the AOHDL-SRP model involves the min–max normalization technique to normalize the input data to a uniform format. Besides, the presented AOHDL-SRP model applies HDL using a convolutional neural network with an attention-oriented long short-term memory (ALSTM) model. Finally, the arithmetic optimization algorithm (AOA) is applied for the hyperparameter optimization of the HDL model, and it assists in improving the predictive performance. The experimental validation of the AOHDL-SRP model is tested, and the results are lower than different expectations. The AOHDL-SRP model has demonstrated that it is superior to all other models by achieving the highest possible R2-score of 100% and MSE, RMSE, and MAE values of 0.18, 0.43, and 0.32, respectively.

The applicability of sunshine-based global solar radiation models modified with meteorological factors for different climate zones of China

With the development of renewable energy, the exploitation and utilization of solar energy resources also need continuous progress, but solar radiation data shortage has become a serious concern. A method for estimating global solar radiation has been developed to address this issue. The sunshine-based model is currently the most widely used model due to its high calculation accuracy and few input parameters. This paper will first review 13 subcategories (8 categories in total) of the global solar radiation prediction model based on sunshine. Subsequently, the astronomical factors were introduced to modify empirical coefficients, and 8 new categories of models based on sunshine rate were introduced. The radiation data from 83 meteorological stations in China was used to train and validate the model, and the performance of the model was evaluated by using evaluation indicators, such as coefficient of determination (R2), root mean square error (RMSE), mean absolute bias error (MABE), mean bias error (MBE), and global performance index (GPI). The results show that the R2 value of the unmodified empirical model is in a range of 0.82–0.99, and the RMSE value is in a range of 0.018–3.09. In contrast, with the introduction of the astronomical factor, the model accuracy improves significantly, and the modified power function model (N3) gains its best performance. The R2 of model N3 is in a range of 0.86–0.99, and the RMSE value is in a range of 0.018–2.62. The R2 increases by 0.49%, while the RMSE value 6.44%. Above all, it does not require the input of other meteorological parameters for predicting the value of global solar radiation.

An evolutionary robust solar radiation prediction model based on WT-CEEMDAN and IASO-optimized outlier robust extreme learning machine

As a kind of clean energy, solar energy occupies a pivotal position in energy applications. Accurate and reliable solar radiation prediction is critical to the application of solar energy. In particular, a novel solar radiation prediction based on wavelet transform (WT), complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN), improved atom search optimization (IASO) and outlier-robust extreme learning machine (ORELM) is proposed for solar radiation prediction. First, WT is used to denoise the original solar radiation time series, and CEEMDAN method is used to decompose the denoised sequence into intrinsic mode function (IMF) components with different distributions according to the fluctuation scale. Then the IASO algorithm is used to optimize the weights and thresholds of the ORELM to improve the performance of the ORELM model. Levy flight is added to the ASO to enhance the local and global search capability while the chaos population initialization based on piecewise linear chaotic map (PWLCM) is taken to improve the randomness and ergodicity of the initial population within the feasible region. Finally, the comparison with other benchmark models verifies the robustness and accuracy of the proposed solar radiation prediction model.

Using of heat thermal storage of PCM and solar energy for distributed clean building heating: A multi-level scale-up research

Solar energy coupled with electric heat storage is a promising energy saving technology for distributed building heating. Energy saving performance of this technology used in buildings has been widely investigated by prototype-scale experiments and numerical assessments. However, the commercial and economic feasibility of this technology during the operational phase of a real building have not been proven yet. This study conducted a comprehensive techno-economic analysis of the electric heat storage system coupled with solar energy installed in a three-story office building (2000 m2 heating area) in Tianjin, China. The electric heat storage system with the heat storage capacity of 1274.8 MJ was installed on the first floor of the building. The solar collector system with a total heat collection area of 160 m2 was installed on the roof of the building. Meanwhile, the long short-term memory was used to predict the heat gain of solar collector in order to adjust the quantity of storage heat of storage device. The results indicated that the prepared storage medium possessed an excellent thermal property with a total heat storage capacity of 448.8 kJ/kg within 50–150 °C. The average thermal efficiency of vacuum tube solar collector was 51.3%, its heat gain per square meter per day was 6.5 MJ. The use of the solar radiation prediction model stabilized thermal efficiency of heat storage reservoir at 95.8%. The specific operational cost of this system within an entire heating season was 15.6 renminbi (RMB)/m2/season. Distribute clean building heating (DCBH) system can save up to 61% of heating cost compared to the centralized heating. The results showed a great application potential of combination of electric heat storage and solar energy based on solar energy prediction for the heat dispatch of distributed building heating.

Solar Radiation Prediction Model for the Yellow River Basin with Deep Learning

Solar radiation is the main source of energy on the Earth’s surface. It is very important for the environment and ecology, water cycle and crop growth. Therefore, it is very important to obtain accurate solar radiation data. In this study, we use the highest temperature Tmax, lowest temperature Tmin, average temperature Tavg, wind speed U, relative humidity RH, sunshine duration H and maximum sunshine duration Hmax as input variables to construct a deep learning prediction model of solar radiation in the Yellow River Basin. It is compared with the recommended and corrected values of the widely used Å-P method. The results show that: (1) The correction results of the Å-P equation are better in the upstream and downstream of the Yellow River Basin but worse in the midstream. (2) The prediction result of the deep learning model in the Yellow River Basin is far better than that of the Å-P equation using the FAO-56 recommended value. It is the best in the downstream of the Yellow River Basin: R2 increases from 0.894 to 0.934; MSE, RMSE and MAE decrease by 43.12%, 27.73% and 25.80%, respectively. The upstream prediction result comes in second: R2 increases from 0.888 to 0.921; MSE, RMSE and MAE decrease by 33.27%, 20.02% and 19.04%, respectively. The midstream result is the worst: R2 increases from 0.869 to 0.874; MSE, RMSE and MAE decrease by −0.50%, 0.07% and 3.82%, respectively. (3) The prediction results of the deep learning model in the upstream and downstream of the Yellow River Basin are far better than those of the Å-P equation using correction. The R2 in the upstream of the Yellow River Basin increases from 0.889 to 0.921. MSE, RMSE and MAE decrease by 22.11%, 11.84% and 8.94%, respectively. R2 in the downstream of the Yellow River Basin increases from 0.900 to 0.934, and MSE, RMSE and MAE decrease by 13.21%, 11.40% and 5.55%, respectively. In the midstream of the Yellow River Basin, the prediction results of the deep learning model are worse than those of the Å-P equation using correction: R2 increases from 0.870 to 0.874, but MSE, RMSE and MAE decrease by −24.93%, −10.83% and −11.56%, respectively.

Solar Radiation Prediction using Ant Colony Optimization and Artificial Neural Network

This paper proposes a Solar Radiation Prediction Model employing Ant Colony Optimization (ACO) and Artificial Neural Network (ANN), named as SRPM. SRPM aims to incorporate the Feature Selection (FS) technique in the ANN training with the guidance of hybridizing ACO. The main reason behind using FS technique is, it can provide an improved solution from a particular problem by identifying the most salient features from the available feature set. In SRPM, the hybridizing ACO search technique utilizes the information gathered from the correlation among the features and the result of ANN training. To assist the ACO search, pheromone updating technique and measurement of heuristic information have been performed by two particular sets of rules. Thus, SRPM utilizes the benefits of using the wrapper and filter approaches in selecting the salient features during SRP task. The combination eventually creates an equipoise between exploration and exploitation of ants in the way of searching as well as strengthening the capability of global search of ACO for obtaining well qualified solution in SRPM. To evaluate the executive efficiency of SRPM, data samples were collected from BMD and NASA-SSE department. Exploratory outcomes show that SRP can select six utmost salient features by providing 99.74% and 99.78% averaged testing accuracies for BMD and NASA-SSE data samples that are composed of 12 and 15 original features, respectively. Furthermore, the proposed model successfully obtained 0.26% and 0.22% MAPE for BMD and NASA-SSE data samples, respectively with a high correlation of about 99.97% within the actual and predicted data.

Improved Complete Ensemble Empirical Mode Decomposition with Adaptive Noise Deep Residual model for short-term multi-step solar radiation prediction

Global Solar Radiation (GSR) prediction models are critical to improve the dispatch, control, and stabilization of solar renewable power, and to integrate the solar energy into the electrical grid system. GSR, especially on a short-term scale, can have important fluctuations, which may affect the total energy expected to be supplied to the grid. To overcome this issue, prediction models with a high forecasting performance are needed. In this paper a novel framework based on Improved Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (ICEEMDAN) and Deep Residual Network with Bidirectional long short-term memory, i.e., DRESNET model, is proposed for obtaining accurate multi-step ahead GSR predictions. To train the proposed ICEEMDAN-DRESNET hybrid model, minute-level daylight data from Energy Sector Management Assistance Program in Nepalgunj (mid-western Nepal) are used. The results demonstrate ICEEMDAN-DRESNET model is an excellent tool for short-term solar energy monitoring, yielding excellent predictions, in all metrics such as MAE 9.769 W/m2, MAPE 5.657%, TIC 1.143, CPI 4.739, and TIC 0.023 for 5-min time-horizon predictions, improving the results from the benchmark models. As the forecasting time-horizon is increased, the ICEEMDAN-DRESNET model accuracy drops, with MAE 33.672 W/m2; MAPE 31.749% for 1-hr, MAE 22.625 W/m2; MAPE 18.312% (30-min) and MAE 14.897 W/m2; MAPE 10.358% (15-min), also better than the benchmark models. The results confirm the competitive merit of ICEEMDAN and DRESNET integration to improve deep learning and the potential of proposed model for the monitoring of solar or other renewable (e.g., wind or solar) energies.

Multi-step solar irradiation prediction based on weather forecast and generative deep learning model

With the rapid development of computer technology, more and more deep learning models are used in solar radiation (irradiation) prediction. There have been a lot of studies discussing the research of this type of model. However, how to better apply the deep learning model in the optimization method of building energy system, such as multi-step solar radiation (irradiation) prediction model in model predictive control (MPC), is still a challenging issue due to the complexity of the time series and the accumulation of errors in multi-step forecasts.In this research, a deep generative model based on LSTM is developed for multi-step solar irradiation prediction at least 24 h in the future. Measured data and temperature forecast data from the Tokyo Meteorological Agency were used for training and testing in this experiment. The results show that generating the model first can effectively avoid the problem of error accumulation. The generative model can produce an accuracy improvement of 7.7% against traditional regression LSTM model. Secondly, the introduction of the temperature forecast data from the previous one day can increase the forecast accuracy by about 18% points. When the earlier temperature forecast is used, the forecast accuracy will gradually decrease, and the use of the temperature forecast released 3 days before can hardly improve the forecast effect. In the end, using hourly temperature forecasts will result in better forecast accuracy than using daily temperature forecasts.

Solar radiation and solar energy estimation using ANN and Fuzzy logic concept: A comprehensive and systematic study.

To overcome the need of the world for energy consumption, we have to find some better and stable alternate ways of renewable energy with advanced technology. The most readily available source of energy is solar energy but solar energy has nonlinear nature due to the random nature of climate conditions. So, one way to solve is solar radiation prediction and solar energy prediction using more accurate techniques. Also, energy business and power system control units require more accuracy along with very short to large duration prediction in advance. So, to complete the requirement many prediction techniques are used and among them, Artificial Neural Network (ANN) and Fuzzy are more accurate and reliable techniques. In this paper basically, a literature study for solar radiation and energy prediction using ANN and Fuzzy logic techniques has been carried out. Many studies are reviewed and then selected some most accurate, reliable, and relevant studies for further study. ANN models with different algorithms such as feed-forward back-propagation-based ANN, Multi-layer feed-forward-based ANN model, Linear regression with ANN model, GNN-based model are reviewed in the study. ANN models with different input parameters combinations and the different number of neurons were also reviewed. Fuzzy logic-based and Adaptive Neuro-Fuzzy interface (ANFIS)-based different models have been reviewed and observed that the ANFIS technique performs better. From the study, it has been noted that ANN and Fuzzy logic employed models are most effective for estimation than any other empirical models. It is found that solar radiation and energy prediction models are dependent on input parameters more. At last, highlighted some possible research opportunities and areas for better efficiency of the results.

Modeling of solar photovoltaic power using a two-stage forecasting system with operation and weather parameters

ABSTRACT The integration of solar photovoltaic (SPV) system to the grid has introduced a new source of intermittency in the grid, and the grid has to react smartly to the changes that occur in the penetration of SPV power. Accurate modeling of weather-dependent SPV power will be helpful in forecasting the penetration of SPV power into the grid. An SPV power output forecasting model has been developed based on artificial neural network (ANN) approach. Two forecasters, namely ANN forecaster and two-stage hybrid-ANN forecaster, are developed with operational and weather parameters. The historical data of SPV power (P), hours of operation of SPV system (to), daily global solar radiation (H), and ambient temperature(T) are used as modeling parameters. The combination of modeling parameters {P, H, T, to} is identified as the best combination that influences the forecasting of day-ahead power output. A relative root mean square error (RRMSE) of 5.74% was obtained with the combination of {P, H, T, to}. An RRMSE of 6.04% was observed with the combination of {P, H, T} as inputs, and the hours of operation of the SPV plant could be ignored in the model. The historical power data of the SPV plant is identified as the crucial parameter in the SPV power forecast model and has given an RRMSE of 7.25%. The models developed with temperature and radiation as modeling parameters have resulted in good forecasting accuracy, which could be best suitable for feasibility studies of SPV plant at a particular location. Solar radiation prediction models are used in the development of hybrid-ANN forecaster. It has produced an RRMSE of 7.35% with four inputs. The hybrid-ANN forecaster with predicted radiations as modeling input will eliminate the need of a costly pyranometer. The models developed in the present study have utilized readily available parameters as modeling parameters, thereby cost of the forecasting system has been decreased. The developed models will be useful for energy scheduling and energy management in the smart grid. Abbreviations: GSR, Global Solar Radiation; ANN, Artificial Neural networks; SPV, Solar Photovoltaics; MAPE, Mean Absolute Percentage Error; RRMSE, Relative Root Mean Square Error; MSE, Mean Square Error.

Forecasting of solar radiation in photovoltaic power station based on ground‐based cloud images and BP neural network

The solar radiation near the surface is the main reason that affects photovoltaic power generation. Accurate ultra-short-term solar radiation prediction is the premise of photovoltaic power generation prediction. Here the cloud movement prediction method based on the ground-based cloud images is presented. The cloud recognition, cloud matching, cloud area correction and cloud movement prediction are performed to predict the drift trajectory of the clouds that will block the sun. Then, using digital image technology, 13 kinds of feature information are extracted from the ground-based cloud images. Then, these feature information are input into BP neural network, and the parameters of BP neural network are optimized by genetic algorithm. Through a large number of data training, a new ultra-short-term prediction model of solar radiation is established. Finally, through experimental comparison, the results show that the prediction accuracy of the model with the feature information of ground-based cloud images can reach 96%, compared with the model without the feature information of ground-based cloud images, the accuracy is improved by 5%. The proposed ultra-short-term solar radiation prediction model can effectively predict the radiation jumping process caused by cloud occlusion, and greatly improve the prediction accuracy, especially in cloudy weather.