Study of hourly and daily solar irradiation forecast using diagonal recurrent wavelet neural networks

An accurate forecast of solar irradiation is required for various solar energy applications and environmental impact analyses in recent years. Comparatively, various irradiation forecast models based on artificial neural networks (ANN) perform much better in accuracy than many conventional prediction models. However, the forecast precision of most existing ANN based forecast models has not been satisfactory to researchers and engineers so far, and the generalization capability of these networks needs further improving. Combining the prominent dynamic properties of a recurrent neural network (RNN) with the enhanced ability of a wavelet neural network (WNN) in mapping nonlinear functions, a diagonal recurrent wavelet neural network (DRWNN) is newly established in this paper to perform fine forecasting of hourly and daily global solar irradiance. Some additional steps, e.g. applying historical information of cloud cover to sample data sets and the cloud cover from the weather forecast to network input, are adopted to help enhance the forecast precision. Besides, a specially scheduled two phase training algorithm is adopted. As examples, both hourly and daily irradiance forecasts are completed using sample data sets in Shanghai and Macau, and comparisons between irradiation models show that the DRWNN models are definitely more accurate.     

BP神经网络模型结构对漫湾径流预报精度的影响研究

以云南省漫湾水电站历史径流状况为研究对象，运用三层前馈反向传播神经网络模型对径流进行中长期预报。为解决神经网络预报模型结构难以确定的问题，尝试在预报过程中通过改变该网络模型的结构并对得到的结果进行比较，从而找到适合该径流序列的最佳神经网络模型结构。实际应用表明，使用该结构的模型在实际预报过程中取得了良好的效果。     

Wind power forecast using neural networks: Tuning with optimization techniques and error analysis

The increased integration of wind power into the power system implies many challenges to the network operators, mainly due to the hard to predict and variability of wind power generation. Thus, an accurate wind power forecast is imperative for systems operators, aiming at an efficient and economical wind power operation and integration into the power system. This work addresses the issue of forecasting short‐term wind speed and wind power for 1 hour ahead, combining artificial neural networks (ANNs) with optimization techniques on real historical wind speed and wind power data. Levenberg‐Marquardt (LM) and particle swarm optimization (PSO) are used as training algorithms to update the weights and bias of the ANN applied to wind speed predictions. The forecasting performance produced by the proposed models are compared with each other, as well as with the benchmark persistence model. Test results show higher performance for ANN‐LM wind speed forecasting model, outperforming both ANN‐PSO and persistence. The application of ANN‐LM to wind power forecast revealed also a good performance, with an average improvement of 2.8% in relation to persistence. An innovative analysis of mean absolute percentage error (MAPE) behaviour in time and in typical days is finally offered in the paper.     

应用小波-人工神经网络组合模型研究电力负荷预报

针对负荷时间序列的非线性和多时间尺度特性．提出了将小波分析与人工神经网络相结合进行负荷预报的方法——小波-人工神经网络组合模型。该模型吸取了小波分析的多分辨功能和人工神经网络的非线性逼近能力。以月、日平均负荷预报为例对模型进行验证．结果表明：该模型的拟合、检验精度较高。     

Use of artificial neural networks for transport energy demand modeling

The paper illustrates an artificial neural network (ANN) approach based on supervised neural networks for the transport energy demand forecasting using socio-economic and transport related indicators. The ANN transport energy demand model is developed. The actual forecast is obtained using a feed forward neural network, trained with back propagation algorithm. In order to investigate the influence of socio-economic indicators on the transport energy demand, the ANN is analyzed based on gross national product (GNP), population and the total annual average veh-km along with historical energy data available from 1970 to 2001. Comparing model predictions with energy data in testing period performs the model validation. The projections are made with two scenarios. It is obtained that the ANN reflects the fluctuation in historical data for both dependent and independent variables. The results obtained bear out the suitability of the adopted methodology for the transport energy-forecasting problem.     

Modeling and forecasting of Turkey’s energy consumption using socio-economic and demographic variables

This study deals with the modeling of the energy consumption in Turkey in order to forecast future projections based on socio-economic and demographic variables (gross domestic product-GDP, population, import and export amounts, and employment) using artificial neural network (ANN) and regression analyses. For this purpose, four diverse models including different indicators were used in the analyses. As the result of the analyses, this research proposes Model 2 as a suitable ANN model (having four independent variables being GDP, population, the amount of import and export) to efficiently estimate the energy consumption for Turkey. The proposed model predicted the energy consumption better than the regression models and the other three ANN models. Thus, the future energy consumption of Turkey is calculated by means of this model under different scenarios. The predicted forecast results by ANN were compared with the official forecasts. Finally, it was concluded that all the scenarios that were analyzed gave lower estimates of the energy consumption than the MENR projections and these scenarios also showed that the future energy consumption of Turkey would vary between 117.0 and 175.4 Mtoe in 2014.     

Hybrid methodology for hourly global radiation forecasting in Mediterranean area

The renewable energies prediction and particularly global radiation forecasting is a challenge studied by a growing number of research teams. This paper proposes an original technique to model the insolation time series based on combining Artificial Neural Network (ANN) and Auto-Regressive and Moving Average (ARMA) model. While ANN by its non-linear nature is effective to predict cloudy days, ARMA techniques are more dedicated to sunny days without cloud occurrences. Thus, three hybrids models are suggested: the first proposes simply to use ARMA for 6 months in spring and summer and to use an optimized ANN for the other part of the year; the second model is equivalent to the first but with a seasonal learning; the last model depends on the error occurred the previous hour. These models were used to forecast the hourly global radiation for five places in Mediterranean area. The forecasting performance was compared among several models: the 3 above mentioned models, the best ANN and ARMA for each location. In the best configuration, the coupling of ANN and ARMA allows an improvement of more than 1%, with a maximum in autumn (3.4%) and a minimum in winter (0.9%) where ANN alone is the best.     

Artificial neural network analysis of Moroccan solar potential

An artificial neural network (ANN) model is used to forecast the annual and monthly solar irradiation in Morocco. Solar irradiation data are taken from the new Satellite Application Facility on Climate Monitoring (CM-SAF)-PVGIS database. The database represents a total of 12 years of data from 1998 to 2010. In this paper, the data are inferred using an ANN algorithm to establish a forward/reverse correspondence between the longitude, latitude, elevation and solar irradiation. Specifically, for the ANN model, a three-layered, back-propagation standard ANN classifier is considered consisting of three layers: input, hidden and output layer. The learning set consists of the normalised longitude, latitude, elevation and the normalised mean annual and monthly solar irradiation of 41 Moroccan sites. The testing set consists of patterns just represented by the input component, while the output component is left unknown and its value results from the ANN algorithm for that specific input. The results are given in the form of the annual and monthly maps. They indicate that the method could be used by researchers or engineers to provide helpful information for decision makers in terms of sites selection, design and planning of new solar plants.     

输入模式和遗传算子对神经网络径流预报精度的影响

采用均匀设计的方法安排数值试验，比较不同的遗传算子组合及数据输入模式对神经网络径流预报精度的影响。研究发现，与未归一数据输入模式相比，归一化数据输入模式使网络预报的精度明显提高；不同算子组合优化神经网络初始权重径流预报精度差别较大，未归一网络的优化效果较归一网络好；同时采用数据归一输入模式与遗传算法优化神经网络初始权重未产生优化效果叠加。     

基于遗传程序设计的中长期径流预报模型研究与应用

应用遗传程序设计建立径流中长期预报模型,结合径流序列数据的特点通过自相关分析确定其滞时输入变量的个数,采用均方误差作为其适应度评价函数,以漫湾实测月径流序列(1953~2003年)和洪家渡实测月径流序列(1951~2004年)为例,通过与ARMA模型、人工神经网络模型的预报结果比较,显示该模型应用于径流中长期预报简单易行且精度较高。     

Assessment of the benefits of numerical weather predictions in wind power forecasting based on statistical methods

Several forecast systems based on Artificial Neural Networks have been developed to predict power production of a wind farm located in a complex terrain, where geographical effects make wind speed predictions difficult) in different time horizons: 1,3,6,12 and 24 h.In the first system, the neural network has been used only as a statistic model based on time series of wind power; later it has been integrated with numerical weather predictions, by which an interesting improvement of the performance has been reached, especially with the longer time horizons. In particular, a sensitivity analysis has been carried out in order to find those numerical weather parameters with the best impact on the forecast.Then, after the implementation of forecast systems based on a single ANN, the two best prediction systems individuated through the sensitivity analysis, have been employed in a hybrid approach, made up of three different ANNs.Besides, a prediction system based on the wavelet decomposition technique has been also carried out in order to evaluate its contribute on the forecast performance in two time horizons (1 and 24 h).The error of the different forecast systems is investigated and the statistical distributions of the error are calculated and presented.     

Short‐term load forecast using trend information and process reconstruction

The algorithms for short‐term load forecast (STLF), especially within the next‐hour horizon, belong to a group of methodologies that aim to render more effective the actions of planning, operating and controlling electric energy systems (EES). In the context of the progressive liberalization of the electricity sector, unbundling of the previous monopolistic structure emphasizes the need for load forecast, particularly at the network level. Methodologies such as artificial neural networks (ANN) have been widely used in next‐hour load forecast. Designing an ANN requires the proper choice of input variables, avoiding overfitting and an unnecessarily complex input vector (IV). This may be achieved by trying to reduce the arbitrariness in the choice of endogenous variables. At a first stage, we have applied the mathematical techniques of process‐reconstruction to the underlying stochastic process, using coding and block entropies to characterize the measure and memory range. At a second stage, the concept of consumption trend in homologous days of previous weeks has been used. The possibility to include weather‐related variables in the IV has also been analysed, the option finally being to establish a model of the non‐weather sensitive type. The paper uses a real‐life case study. Copyright © 2006 John Wiley & Sons, Ltd.     

Forecasting energy consumption in Taiwan using hybrid nonlinear models

The total consumption of electricity and petroleum energies accounts for almost 90% of the total energy consumption in Taiwan, so it is critical to model and forecast them accurately. For univariate modeling, this paper proposes two new hybrid nonlinear models that combine a linear model with an artificial neural network (ANN) to develop adjusted forecasts, taking into account heteroscedasticity in the model's input. Both of the hybrid models can decrease round-off and prediction errors for multi-step-ahead forecasting. The results suggest that the new hybrid model generally produces forecasts which, on the basis of out-of-sample forecast encompassing tests and comparisons of three different statistic measures, routinely dominate the forecasts from conventional linear models. The superiority of the hybrid ANNs is due to their flexibility to account for potentially complex nonlinear relationships that are not easily captured by linear models. Furthermore, all of the linear and nonlinear models have highly accurate forecasts, since the mean absolute percentage forecast error (MAPE) results are less than 5%. Overall, the inclusion of heteroscedastic variations in the input layer of the hybrid univariate model could help improve the modeling accuracy for multi-step-ahead forecasting.     

Hybrid intra-hour DNI forecasts with sky image processing enhanced by stochastic learning

We propose novel smart forecasting models for Direct Normal Irradiance (DNI) that combine sky image processing with Artificial Neural Network (ANN) optimization schemes. The forecasting models, which were developed for over 6 months of intra-minute imaging and irradiance measurements, are used to predict 1 min average DNI for specific time horizons of 5 and 10 min. We discuss optimal models for low and high DNI variability seasons. The different methods used to develop these season-specific models consist of sky image processing, deterministic and ANN forecasting models, a genetic algorithm (GA) overseeing model optimization and two alternative methods for training and validation. The validation process is carried over by the Cross Validation Method (CVM) and by a randomized training and validation set method (RTM). The forecast performance for each solar variability season is evaluated, and the models with the best forecasting skill for each season are selected to build a hybrid model that exhibits optimal performance for all seasons. An independent testing set is used to assess the performance of all forecasting models. Performance is assessed in terms of common error statistics (mean bias and root mean square error), but also in terms of forecasting skill over persistence. The hybrid forecast models proposed in this work achieve statistically robust forecasting skills in excess of 20% over persistence for both 5 and 10 min ahead forecasts, respectively.     

基于EEMD-ANN的水库年径流预测

以三峡水库为例,基于集合经验模态分解(EEMD)及人工神经网络方法对水库年径流进行预测。首先利用Mann-Kendall和Pettitt法对水库年径流序列进行突变检测,获得平稳径流序列,然后采取EEMD方法分解径流序列,得到固有模态函数(IMF)和残差,最后对不同IMFs和残差分别建立人工神经网络预测模型,叠加所有模型预测结果得到年径流预测值。结果表明,基于EEMD-ANN的年径流预测模型优于自回归模型和人工神经网络模型,其预测结果与实测值的相关性更强,预测误差分别减少了11.4%、8.7%。同时,构建EEMD-ANN预测模型时需考虑径流序列的突变特征,采取平稳径流序列的预测效果更优。     

Optimization of an artificial neural network dedicated to the multivariate forecasting of daily global radiation

This paper presents an application of Artificial Neural Networks (ANNs) to predict daily solar radiation. We look at the Multi-Layer Perceptron (MLP) network which is the most used of ANNs architectures. In previous studies, we have developed an ad-hoc time series preprocessing and optimized a MLP with endogenous inputs in order to forecast the solar radiation on a horizontal surface. We propose in this paper to study the contribution of exogenous meteorological data (multivariate method) as time series to our optimized MLP and compare with different forecasting methods: a naïve forecaster (persistence), ARIMA reference predictor, an ANN with preprocessing using only endogenous inputs (univariate method) and an ANN with preprocessing using endogenous and exogenous inputs. The use of exogenous data generates an nRMSE decrease between 0.5% and 1% for two stations during 2006 and 2007 (Corsica Island, France). The prediction results are also relevant for the concrete case of a tilted PV wall (1.175 kWp). The addition of endogenous and exogenous data allows a 1% decrease of the nRMSE over a 6 months-cloudy period for the power production. While the use of exogenous data shows an interest in winter, endogenous data as inputs on a preprocessed ANN seem sufficient in summer.     

Forecasting annual gross electricity demand by artificial neural networks using predicted values of socio-economic indicators and climatic conditions: Case of Turkey

In this work, the annual gross electricity demand of Turkey was modeled by multiple linear regression and artificial neural networks as a function population, gross domestic product per capita, inflation percentage, unemployment percentage, average summer temperature and average winter temperature. Among these, the unemployment percentage and the average winter temperature were found to be insignificant to determine the demand for the years between 1975 and 2013. Next, the future values of the statistically significant variables were predicted by time series ANN models, and these were simulated in a multilayer perceptron ANN model to forecast the future annual electricity demand. The results were validated with a very high accuracy for the years that the electricity demand was known (2007–2013), and they were also superior to the official predictions (done by Ministry of Energy and Natural Resources of Turkey). The model was then used to forecast the annual gross electricity demand for the future years, and it was found that, the demand will be doubled reaching about 460 TW h in the year 2028. Finally, it was concluded that the approach applied in this work can easily be implemented for other countries to make accurate predictions for the future.     

基于自适应模糊推理系统模型的径流中长期预报

介绍了自适应模糊推理系统ANFIS的原理结构及学习算法。以漫湾和双牌两座水库实测月径流序列为研究对象，研究不同的输入及不同的模糊数对自适应模糊推理系统模型做中长期预报的影响，并通过与人工神经网络模型的预报结果进行比较，显示本模型是中长期水文预报方法中较为准确的方法之一。     

基于人工神经网络模型的风速预测

利用在达坂城风电场30m轮毂高处的1min实测风速数据,采用人工神经网络模型ANN对未来短时间风速进行预报。通过对风速反复训练与检测来选择一组合适的模型参数,并对模型进行了误差分析。研究结果表明,使用BP神经网络对未来风速进行短时间预测能够达到较好的效果,误差较ARMA模型更精确,但是对于突变信息的处理能力仍然有限。     

Forecasting of peak electricity demand in Mauritius using the non-homogeneous Gompertz diffusion process

In this study, the non-homogeneous Gompertz diffusion process (NHGDP) is used to model the monthly peak electricity demand in Mauritius in order to predict the future values on the basis of a Genetic Algorithm (GA) approach. Our model is developed based a key economic indicator which is the gross domestic product (GDP) and the weather factors such as temperature, hours of sunshine and humidity. Genetic Algorithm then searches for the best coefficients by minimizing the root mean square error. Monthly data from January 2005 to December 2008 are considered to test the model. Finally, the Artificial Neural Network (ANN) is used to forecast each independent variable for the year 2009 and the NHGDP model is validated for that year. Our results show that the model provides an accurate and reliable prediction for the monthly peak electricity demand in Mauritius.