An ANN-based approach for predicting global radiation in locations with no direct measurement instrumentation

In this work, a novel approach using an artificial neural network was used to develop a model for analyzing the relationship between the Global Radiation (GR) and climatological variables, and to predict GR for locations not covered by the model's training data. The predicted global radiation values for the different locations (for different months) were then compared with the actual values. Results indicate that the model predicted the Global Radiation values with a good accuracy of approximately 93% and a mean absolute percentage error of 7.30. In addition, the model was also tested to predict GR values for the Seeb location over a 12 month period. The monthly predicted values of the ANN model compared to the actual GR values for Seeb produced an accuracy of 95% and a mean absolute percentage error of 5.43. Data for these locations were not included as part of ANN training data. Hence, these results demonstrate the generalization capability of this novel approach over unseen data and its ability to produce accurate estimates. Finally, this ANN-based model was also used to predict the global radiation values for Majees, a new location in north Oman.     

Prediction of monthly mean daily diffuse solar radiation using artificial neural networks and comparison with other empirical models

An artificial neural network (ANN) model for estimating monthly mean daily diffuse solar radiation is presented in this paper. Solar radiation data from 9 stations having different climatic conditions all over China during 1995–2004 are used for training and testing the ANN. Solar radiation data from eight typical cities are used for training the neural networks and data from the remaining one location are used for testing the estimated values. Estimated values are compared with measured values in terms of mean percentage error (MPE), mean bias error (MBE) and root mean square error (RMSE). The results of the ANN model have been compared with other empirical regression models. The solar radiation estimations by ANN are in good agreement with the actual values and are superior to those of other available models. In addition, ANN model is tested to predict the same components for Zhengzhou station over the same period. Results indicate that ANN model predicts the actual values for Zhengzhou with a good accuracy of 94.81%. Data for Zhengzhou are not included as a part of ANN training set. Hence, these results demonstrate the generalization capability of this approach and its ability to produce accurate estimates in China.     

Long-term validated simulation of a building integrated photovoltaic system

Electrical and thermal simulations of a building integrated photovoltaic system were undertaken with a transient system simulation program using real field input weather data. Predicted results were compared with actual measured data. A site dependent global-diffuse correlation is proposed. The best-tilted surface radiation model for estimating insolation on the inclined surface was selected by statistical tests. To predict the module temperature, a linear correlation equation is developed which relates the temperature difference between module and ambient to insolation. Different combinations of tilted surface radiation model, global-diffuse correlation model and predicted module temperature were used to carry out the simulation and corresponding simulated results compared with the measured data to determine the best combination which gave the least error. Results show that modification of global-diffuse correlation and module temperature prediction improved the overall accuracy of the simulation model. The monthly error between measured and predicted PV output was lied below 16%. Over the period of simulation, the monthly average error between measured and predicted PV output was estimated to be 6.79% whereas, the monthly average error between measured and predicted inverter output was 4.74%.     

Modeling solar still production using local weather data and artificial neural networks

A study has been performed to predict solar still distillate production from single examples of two different commercial solar stills that were operated for a year and a half. The purpose of this study was to determine the effectiveness of modeling solar still distillate production using artificial neural networks (ANNs) and local weather data. The study used the principal weather variables affecting solar still performance, which are the daily total insolation, daily average wind velocity, daily average cloud cover, daily average wind direction and daily average ambient temperature. The objectives of the study were to assess the sensitivity of the ANN predictions to different combinations of input parameters as well as to determine the minimum amount of inputs necessary to accurately model solar still performance. It was found that 31-78% of ANN model predictions were within 10% of the actual yield depending on the input variables that were selected. By using the coefficient of determination, it was found that 93-97% of the variance was accounted for by the ANN model. About one half to two thirds of the available long term input data were needed to have at least 60% of the model predictions fall within 10% of the actual yield. Satisfactory results for two different solar stills suggest that, with sufficient input data, the ANN method could be extended to predict the performance of other solar still designs in different climate regimes.     

Assessment of diffuse solar energy under general sky condition using artificial neural network

In this paper, artificial neural network (ANN) models are developed for estimating monthly mean hourly and daily diffuse solar radiation. Solar radiation data from 10 Indian stations, having different climatic conditions, all over India have been used for training and testing the ANN model. The coefficient of determination (R²) for all the stations are higher than 0.85, indicating strong correlation between diffuse solar radiation and selected input parameters. The feedforward back-propagation algorithm is used in this analysis. Results of ANN models have been compared with the measured data on the basis of percentage root-mean-square error (RMSE) and mean bias error (MBE). It is found that maximum value of RMSE in ANN model is 8.8% (Vishakhapatnam, September) in the prediction of hourly diffuse solar radiation. However, for other stations same error is less than 5.1%. The computation of monthly mean daily diffuse solar radiation is also carried out and the results so obtained have been compared with those of other empirical models. The ANN model shows the maximum RMSE of 4.5% for daily diffuse radiation, while for other empirical models the same error is 37.4%. This shows that ANN model is more accurate and versatile as compared to other models to predict hourly and daily diffuse solar radiation.     

Generation of synthetic daily global solar radiation data based on ERA-Interim reanalysis and artificial neural networks

Four variables (total cloud cover, skin temperature, total column water vapour and total column ozone) from meteorological reanalysis were used to generate synthetic daily global solar radiation via artificial neural network (ANN) techniques. The goal of our study was to predict solar radiation values in locations without ground measurements, by using the reanalysis data as an alternative to the use of satellite imagery. The model was validated in Andalusia (Spain), using measured data for nine years from 83 ground stations spread over the region. The geographical location (latitude, longitude), the day of the year, the daily clear sky global radiation, and the four meteorological variables were used as input data, while the daily global solar radiation was the only output of the ANN. Sixty five ground stations were used as training dataset and eighteen stations as independent dataset. The optimum network architecture yielded a root mean square error of 16.4% and a correlation coefficient of 94% for the testing stations. Furthermore, we have successfully tested the forecasting capability of the model with measured radiation values at a later time. These results demonstrate the generalization capability of this approach over unseen data and its ability to produce accurate estimates and forecasts.     

利用神经网络估算太阳辐射

太阳辐射是一项对太阳能利用，建筑能耗分析和农业等十分重要的气象数据，本文建立了日总太阳辐射月均值的神经网络估算模型，在此基础上利用北京市1971年至1995年的气象数据资料对神经网络进行了训练，用1996至2000年的数据对神经网络的估算进行了检验，并与其它经验模型的估算结果进行了对比，结果表明神经网络的估算结果与实测值吻合的较好，并且精度高于其它经验模型。因此利用神经网络来估算太阳辐射具有很好的应用前景。     

Measurement of solar energy radiation at Nsukka and the determination of the regression coefficients

Solar radiation for Nsukka, latitude 6.8°N, longitude 7.35°E, located 488 m above sea level, was collected for 11 yr using a Gunn-Bellani instrument, and the data obtained were restandardized using an Eppley precision pyranometer. The sunshine data during this period were also obtained using a Campbell-Stokes sunshine recorder.

It is observed that the insolation level for the month of November during the period of measurement is the highest, with an average of 17.50 MJ/m². The level of insolation during this month varied sinusoidally with an amplitude of 3.84 MJ/m². The clearness index, k_T, is 0.53, and there is an indication that the atmosphere was persistently laden with dust. August has the least insolation level with an average of 11.86 MJ/m² and a k_T of 0.32. The atmosphere during this month was always covered with cloud. This work confirmed the assertion by Awachie and co-workers that dust and haze attenuate insolation less than cloud cover. The Nsukka weather is rated to be heavily overcast, and over 90% of the total solar radiation is diffuse, with an average k_T value of 0.43.

The average regression coefficients a and b for Nsukka are 0.21 and 0.51 respectively. These values do not agree with the general relations and values already quoted by some workers. The predicted insolation values for Nsukka using these coefficients in the Angstrom type of formula agree with the measured data with an error of 0.7%. This level of accuracy compares well with those obtained when the insolation values are predicted for each year using the values of a and b deduced for the respective year. Furthermore, there was an indication that the level of accuracy obtained using average values of a and b might increase if a longer period is considered. Thus, with reliable average values of a and b obtained over a reasonably long period, and knowledge of the bright sunshine hours, the measurement of solar radiation in a location, for design purposes, may not be necessary.     

Computation of monthly mean daily global solar radiation in China using artificial neural networks and comparison with other empirical models

In this paper, an artificial neural network (ANN) model is developed for estimating monthly mean daily global solar radiation of 8 typical cities in China. The feed-forward back-propagation algorithm is applied in this analysis. The results of the ANN model and other empirical regression models have been compared with measured data on the basis of mean percentage error (MPE), mean bias error (MBE) and root mean square error (RMSE). It is found that the solar radiation estimations by ANN are in good agreement with the measured values and are superior to those of other available empirical models. In addition, ANN model is tested to predict the same components for Kashi, Geermu, Shenyang, Chengdu and Zhengzhou stations over the same period. Data for Kashi, Geermu, Shenyang, Chengdu and Zhengzhou are not used in the training of the networks. Results obtained indicate that the ANN model can successfully be used for the estimation of monthly mean daily global solar radiation for Kashi, Geermu, Shenyang, Chengdu and Zhengzhou. These results testify the generalization capability of the ANN model and its ability to produce accurate estimates in China.     

A simplified model for generating sequences of global solar radiation data for isolated sites: Using artificial neural network and a library of Markov transition matrices approach

The purpose of this work is to develop a hybrid model which will be used to predict the daily global solar radiation data by combining between an artificial neural network (ANN) and a library of Markov transition matrices (MTM) approach. Developed model can generate a sequence of global solar radiation data using a minimum of input data (latitude, longitude and altitude), especially in isolated sites. A data base of daily global solar radiation data has been collected from 60 meteorological stations in Algeria during 1991–2000. Also a typical meteorological year (TMY) has been built from this database. Firstly, a neural network block has been trained based on 60 known monthly solar radiation data from the TMY. In this way, the network was trained to accept and even handle a number of unusual cases. The neural network can generate the monthly solar radiation data. Secondly, these data have been divided by corresponding extraterrestrial value in order to obtain the monthly clearness index values. Based on these monthly clearness indexes and using a library of MTM block we can generate the sequences of daily clearness indexes. Known data were subsequently used to investigate the accuracy of the prediction. Furthermore, the unknown validation data set produced very accurate prediction; with an RMSE error not exceeding 8% between the measured and predicted data. A correlation coefficient ranging from 90% and 92% have been obtained; also this model has been compared to the traditional models AR, ARMA, Markov chain, MTM and measured data. Results obtained indicate that the proposed model can successfully be used for the estimation of the daily solar radiation data for any locations in Algeria by using as input the altitude, the longitude, and the latitude. Also, the model can be generalized for any location in the world. An application of sizing PV systems in isolated sites has been applied in order to confirm the validity of this model.     

Novel approach in developing hourly global solar radiation model based on peak sunshine hours

The present study explores a novel approach to derive the hourly global solar radiation (HGSR) for any given latitude based on the peak sunshine hours (PSHs). The proposed analytical model describes a relationship between the HGSR and the day length and the PSHs. The applicability of this model is evaluated by comparing the actual and derived values of HGSR for two cities Chennai (13°04′N, 80°17′E) and New Delhi (29°06′N, 77°22′E). To judge the goodness of the proposed model a set of error metrics has been developed by evaluating the variation of actual HGSR from the simulated value for a given day over 12 months. The overall average mean bias error for one year is 1.015% and 1.08% for Chennai and New Delhi, respectively. The agreement between the actual and the simulated values is generally good, with an appreciable correlation of 95%. In particular, unlike the other models this approach requires only two inputs which are easily available for any location. The proposed technique is useful for any solar application designer for deriving the hourly solar radiation values for a given day of any location with less available climate data.     

倾斜面太阳辐射量计算方法研究

倾斜面辐射数据是保证准确设计太阳能利用系统的基础数据,一般由水平面数据计算得出。针对现有计算模型误差较大的现状考虑,试验测试了水平面及不同倾角斜面上的太阳辐射数据,提出了针对直接辐射转换系数的修正方法;通过倾斜面散射辐射数据的计算和分析,在散射辐射模型的基础上,建立耦合计算模型。试验与模型计算结果表明,散射辐射的各向同性受天气工况的影响,耦合模型具有较高的准确性,计算值与实测值的偏差可控制在5.3%以内。     

Evaluation and comparison of hourly solar radiation models

In this paper, an attempt has been made to develop a new model to evaluate the hourly solar radiation for composite climate of New Delhi. The comparison of new model for hourly solar radiation has been carried out by using various model proposed by others. The root mean square error (RMSE) and mean bias error (MBE) have been used to compare the accuracy of new and others model. The results show that the ASHRAE and new proposed model estimate hourly solar radiation better for composite climate of New Delhi in comparison to other models. Hourly solar radiation estimated by constants obtained by new model (modified ASHRAE model) for composite climate of India is fairly comparable with measured data. The percentage mean bias error with new constants for New Delhi was found as low as 0.15 and 0% for hourly beam and diffuse radiation, respectively. There is a 1.9–8.5% RMSE between observed and predicted values of beam radiation using new constants for clear days. The statistical analysis has been used for the present study. Copyright © 2008 John Wiley & Sons, Ltd.     

Computation of beam solar radiation at normal incidence using artificial neural network

In this paper, an artificial neural network (ANN) model is developed for estimating beam solar radiation. Introducing a newly defined parameter, known as reference clearness index (RCI), computation of monthly mean daily beam solar radiation at normal incidence has been carried out. This RCI is defined as the ratio of measured beam solar radiation at normal incidence to the beam solar radiation as computed by Hottel's clear day model. Solar radiation data from 11 stations having different climatic conditions all over India have been used for training and testing the ANN. The feedforward back-propagation algorithm is used in this analysis. The results of ANN model have been compared with measured data on the basis of root mean square error (RMSE) and mean bias error (MBE). It is found that RMSE in the ANN model varies 1.65–2.79% for Indian region.     

Solar radiation modelling for the simulation of photovoltaic systems

The effect of 12 different combinations of diffuse–global correlations and tilted surface radiation models on the accuracy of PV output simulation of a grid-connected photovoltaic (PV) system was studied using statistical methods. A site specific diffuse–global correlation was developed using local insolation data and the performance of this model was compared with those of two other diffuse–global correlations. The impact of diffuse–global correlations on the calculated inclined insolation for four different tilted surface radiation models was investigated on annual, seasonal and monthly basis. The accuracies of predicted 45° inclined plane insolation and PV output were improved when the site specific diffuse–global correlation was used in the main simulation model. The error between measured and predicted inclined insolation was higher in winter than summer. The prediction of PV output was improved by using an isotropic sky tilted surface radiation model instead of the anisotropic models. The accuracy of PV output was also improved when the proposed diffuse–global correlation was used.     

Heat transfer analysis of phase change process in a finned-tube thermal energy storage system using artificial neural network

In this study, a feed-forward back-propagation artificial neural network (ANN) algorithm is proposed for heat transfer analysis of phase change process in a finned-tube, latent heat thermal energy storage system. Heat storage through phase change material (PCM) around the finned tube is experimentally studied. A numerical study is performed to investigate the effect of fin and flow parameter by the solving governing equations for the heat transfer fluid, pipe wall and phase change material. Learning process is applied to correlate the total heat stored in different fin types of tubes, various Reynolds numbers and different inlet temperatures. A number of hidden numbers of ANN are trained for the best output prediction of the heat storage. The predicted total heat storage values obtained by an ANN model with extensive sets of non-training experimental data are then compared with experimental measurements and numerical results. The trained ANN model with an absolute mean relative error of 5.58% shows good performance to predict the total amount of heat stored. The ANN results are found to be more accurate than the numerical model results. The present study using ANN approach for heat transfer analysis in phase change heat storage process appears to be significant for practical thermal energy storage applications.     

The all-sky meteorological radiation model: proposed improvements

The all-sky meteorological radiation model is a broadband solar-radiation estimation model that uses synoptic and sunshine information. The original model due to Muneer–Gul–Kambezidis was improved using regressions based on the sunshine fraction to increase the accuracy of the estimation of diffuse horizontal irradiation, thus achieving an accuracy increase for the estimation of the global horizontal irradiation. The improved model was validated using data from ten worldwide sites and using three statistical indicators:-coefficient of determination between computed and measured global irradiation data and the relevant, mean bias error and the root mean square error of the computed global irradiation. The performance of the new model was improved when compared to that of the original model. The new regression coefficients were found to be more accurate in estimating global horizontal radiation for both fine and coarse datasets.     

基于GA-BP神经网络的逐时总辐射分组模型研究

通过分析影响太阳辐射的主要因素,提出以太阳高度角、季节和天气（晴空指数）作为数据划分依据的分组模型建立方法。以拉萨和西安地区的逐时气象数据和辐射数据为例,基于遗传算法（genetic algorithm,GA）优化的BP神经网络,建立太阳高度角、季节和天气类型的逐时总辐射分组模型。该研究揭示分组模型误差变化的规律,并将其估算误差与AllData模型比较。结果显示,相较于AllData模型,分组模型的估算误差均有降低。其中,天气分组模型误差最小,且西安的天气分组模型结果优于拉萨。西安天气分组模型平均绝对百分比误差（MAPE）和相对均方根误差（rRMSE）相较AllData模型结果分别下降3.96%和4.18%。研究结果表明分组模型能够降低逐时总辐射估算误差,可为估算逐时总辐射提供方法借鉴。     

MODELING OF SOLAR DOMESTIC WATER HEATING SYSTEMS USING ARTIFICIAL NEURAL NETWORKS

Artificial Neural Networks (ANN) are widely accepted as a technology offering an alternative way to tackle complex and ill-defined problems. They can be trained to predict results from examples, are fault tolerant, are able to deal with non-linear problems, and once trained can perform prediction at high speed. ANNs have been used in diverse applications and they have shown to be particularly useful in system modeling and for system identification. The objective of this work was to train an ANN to learn to predict the useful energy extracted and the temperature rise in the stored water of solar domestic water heating (SDHW) systems with the minimum of input data. An ANN has been trained based on 30 known cases of systems, varying from collector areas between 1.81 m² and 4.38 m². Open and closed systems have been considered both with horizontal and vertical storage tanks. In addition to the above, an attempt was made to consider a large variety of weather conditions. In this way the network was trained to accept and handle a number of unusual cases. The data presented as input were the collector area, storage tank heat loss coefficient (U-value), tank type, storage volume, type of system, and ten readings from real experiments of total daily solar radiation, mean ambient air temperature, and the water temperature in the storage tank at the beginning of a day. The network output is the useful energy extracted from the system and the temperature rise in the stored water. The statistical R²-value obtained for the training data set was equal to 0.9722 and 0.9751 for the two output parameters respectively. Unknown data were subsequently used to investigate the accuracy of prediction. These include systems considered for the training of the network at different weather conditions and completely unknown systems. Predictions within 7.1% and 9.7% were obtained respectively. These results indicate that the proposed method can successfully be used for the estimation of the useful energy extracted from the system and the temperature rise in the stored water. The advantages of this approach compared to the conventional algorithmic methods are the speed, the simplicity, and the capacity of the network to learn from examples. This is done by embedding experiential knowledge in the network. Additionally, actual weather data have been used for the training of the network, which leads to more realistic results as compared to other modeling programs, which rely on TMY data that are not necessarily similar to the actual environment in which a system operates.     

Evaluation of models to predict insolation on tilted surfaces

An empirical study was performed to evaluate the validity of various insolation models which employ either an isotropic or an anisotropic distribution approximation for sky light when predicting insolation on tilted surfaces. Data sets of measured hourly insolation values were obtained over a 6-month period using pyranometers which received diffuse and total solar radiation on a horizontal plane and total radiation on surfaces tilted toward the equator at 37° and 60° angles above the horizon. Data on the horizontal surfaces were used in the insolation models to predict insolation on the tilted surface; comparisons of measured vs calculated insolation on the tilted surface were examined to test the validity of the sky light approximations. It was found that the Liu-Jordan isotropic distribution model provides a good fit to empirical data under overcast skies but underestimates the amount of solar radiation incident on tilted surfaces under clear and partly cloudy conditions. The anisotropic-clear-sky distribution model by Temps and Coulson provides a good prediction for clear skies but overstimates the solar radiation when used for cloudy days. An anisotropic-all-sky model was formulated in this effort which provided excellent agreement between measured and predicted insolation throughout the 6-month period.