基于PSO-BP神经网络的短期光伏系统发电预测

对光伏发电影响因素进行了分析,建立了粒子群算法优化的前向神经网络光伏系统发电预测模型。该模型利用了粒子群算法来优化神经网络内部连接权值和阈值,兼具粒子群和BP神经模型的优点,具有较好的收敛速度,泛化性能与预测精度。将光伏电站发电历史数据与天气情况作为样本,运用所建立的模型进行了训练与预测。结果表明,经过粒子群优化的BP网络模型预测精度高于典型BP网络,验证了该方法的有效性。     

Online short-term solar power forecasting

This paper describes a new approach to online forecasting of power production from PV systems. The method is suited to online forecasting in many applications and in this paper it is used to predict hourly values of solar power for horizons of up to 36 h. The data used is 15-min observations of solar power from 21 PV systems located on rooftops in a small village in Denmark. The suggested method is a two-stage method where first a statistical normalization of the solar power is obtained using a clear sky model. The clear sky model is found using statistical smoothing techniques. Then forecasts of the normalized solar power are calculated using adaptive linear time series models. Both autoregressive (AR) and AR with exogenous input (ARX) models are evaluated, where the latter takes numerical weather predictions (NWPs) as input. The results indicate that for forecasts up to 2 h ahead the most important input is the available observations of solar power, while for longer horizons NWPs are the most important input. A root mean square error improvement of around 35% is achieved by the ARX model compared to a proposed reference model.     

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.     

Short term solar irradiance forecasting using a mixed wavelet neural network

In modern smart grids and deregulated electricity markets, accurate forecasting of solar irradiance is critical for determining the total energy generated by PV systems. We propose a mixed wavelet neural network (WNN) in this paper for short-term solar irradiance forecasting, with initial application in tropical Singapore. The key advantage of using wavelet transform (WT) based methods is the high signal compression ability of wavelets, making them suitable for modeling of nonstationary environmental parameters with high information content, such as short timescale solar irradiance. In this WNN, a combination of the commonly known Morlet and Mexican hat wavelets is used as the activation function for hidden-layer neurons of a feed forward artificial neural network (ANN). To demonstrate the effectiveness of the proposed approach, hourly predictions of solar irradiance, which is an aggregate sum of irradiance value observed using 25 sensors across Singapore, are considered. The forecasted results show that WNN delivers better prediction skill when compared with other forecasting techniques.     

A regime-dependent artificial neural network technique for short-range solar irradiance forecasting

Solar power can provide substantial power supply to the grid; however, it is also a highly variable energy source due to changes in weather conditions, i.e. clouds, that can cause rapid changes in solar power output. Independent systems operators (ISOs) and regional transmission organizations (RTOs) monitor the demand load and direct power generation from utilities, define operating limits and create contingency plans to balance the load with the available power generation resources. ISOs, RTOs, and utilities will require solar irradiance forecasts to effectively and efficiently balance the energy grid as the penetration of solar power increases. This study presents a cloud regime-dependent short-range solar irradiance forecasting system to provide 15-min average clearness index forecasts for 15-min, 60-min, 120-min and 180-min lead-times. A k-means algorithm identifies the cloud regime based on surface weather observations and irradiance observations. Then, Artificial Neural Networks (ANNs) are trained to predict the clearness index. This regime-dependent system makes a more accurate deterministic forecast than a global ANN or clearness index persistence and produces more accurate predictions of expected irradiance variability than assuming climatological average variability.     

Improvement in artificial neural network-based estimation of grid connected photovoltaic power output

This paper presents a method to improve the accuracy of artificial neural network (ANN)–based estimation of photovoltaic (PV) power output by introducing two more inputs, solar zenith angle and solar azimuth angle, in addition to the most widely used environmental information, plane-of-array irradiance and module temperature. Solar zenith angle and solar azimuth angle define the solar position in the sky; hence, the loss of modeling accuracy due to impacts of solar angle-of-incidence and solar spectrum is reduced or eliminated. The observed data from two sites where local climates are significantly different is used to train and test the proposed network. The good performance of the proposed network is verified by comparing with existing ANN model, algebraic model, and polynomial regression model which use environmental information only of plane-of-array irradiance and module temperature. Our results show that the proposed ANN model greatly improves the accuracy of estimation in the long term under various weather conditions. It is also demonstrated that the improvement in estimating outdoor PV power output by adding solar zenith angle and azimuth angle as inputs is useful for other data-driven methods like support vector machine regression and Gaussian process regression.     

On the investigation of photovoltaic output power reduction due to dust accumulation and weather conditions

Certain environmental conditions such as accumulation of dust and change in weather conditions affect the amount of solar radiation received by photovoltaic (PV) panel surfaces and thus have a significant effect on panel efficiency. This study conducted an experimental investigation in Surabaya, Indonesia, on the effect of these factors on output PV power reduction from the surface of a PV module. The module was exposed to outside weather conditions and connected to a measurement system developed using a rule-based model to identify different environmental conditions. The rule-based model, a clear sky solar irradiance model that included solar position, and a PV temperature model were then used to estimate the PV output power, and tests were also conducted using an ARM Cortex-M4 microcontroller STM32F407 as a standalone digital controller equipped with voltage, current, temperature, and humidity sensors to measure real time PV output power. In this system, humidity was monitored to identify dusty, cloudy, and rainy conditions. Validated test results demonstrate that the prediction error of PV power output based on the model is 3.6% compared to field measurements under clean surface conditions. The effects of dust accumulation and weather conditions on PV panel power output were then analyzed after one to four weeks of exposure. Results revealed that two weeks of dust accumulation caused a PV power output reduction of 10.8% in an average relative humidity of 52.24%. Results of the experiment under rainy conditions revealed a decrease in PV output power of more than 40% in average relative humidity of 76.32%, and a decrease in output power during cloudy conditions of more than 45% in an average relative humidity of 60.45% was observed. This study reveals that local environmental conditions, i.e., dust, rain, and partial cloud, significantly reduce PV power output.     

基于IFOA-GRNN的电力系统短期负荷预测

考虑到电网负荷与诸多因素有关,设计了一种带有温度、气象、日期类型的广义回归神经网络（GRNN）负荷预测模型。为了提高该模型的预测精度,提出了一种改进果蝇优化算法优化广义回归神经网络（IFOA-GRNN）的方法,即在利用果蝇优化算法（FOA）进入迭代寻优时,通过改进搜索距离优化该算法的性能和稳定性。利用改进的FOA优化GRNN的光滑参数,然后利用训练好的预测模型对甘肃省某地区进行了短期负荷预测,并与FOA-GRNN和误差反向传播神经网络(BPNN)模型结果进行了误差比较。结果表明, IFOA-GRNN具有较高的预测精度,能够满足电力系统短期负荷预测的要求。     

Wind turbine power curve modelling using artificial neural network

Technical improvements over the past decade have increased the size and power output capacity of wind power plants. Small increases in power performance are now financially attractive to owners. For this reason, the need for more accurate evaluations of wind turbine power curves is increasing. New investigations are underway with the main objective of improving the precision of power curve modeling. Due to the non-linear relationship between the power output of a turbine and its primary and derived parameters, Artificial Neural Network (ANN) has proven to be well suited for power curve modelling. It has been shown that a multi-stage modelling techniques using multilayer perceptron with two layers of neurons was able to reduce the level of both the absolute and random error in comparison with IEC methods and other newly developed modelling techniques. This newly developed ANN modeling technique also demonstrated its ability to simultaneously handle more than two parameters. Wind turbine power curves with six parameters have been modelled successfully. The choice of the six parameters is crucial and has been selected amongst more than fifty parameters tested in term of variability in differences between observed and predicted power output. Further input parameters could be added as needed.     

A 24-h forecast of solar irradiance using artificial neural network: Application for performance prediction of a grid-connected PV plant at Trieste, Italy

Forecasting of solar irradiance is in general significant for planning the operations of power plants which convert renewable energies into electricity. In particular, the possibility to predict the solar irradiance (up to 24 h or even more) can became - with reference to the Grid Connected Photovoltaic Plants (GCPV) - fundamental in making power dispatching plans and - with reference to stand alone and hybrid systems - also a useful reference for improving the control algorithms of charge controllers. In this paper, a practical method for solar irradiance forecast using artificial neural network (ANN) is presented. The proposed Multilayer Perceptron MLP-model makes it possible to forecast the solar irradiance on a base of 24 h using the present values of the mean daily solar irradiance and air temperature. An experimental database of solar irradiance and air temperature data (from July 1st 2008 to May 23rd 2009 and from November 23rd 2009 to January 24th 2010) has been used. The database has been collected in Trieste (latitude 45°40′N, longitude 13°46′E), Italy. In order to check the generalization capability of the MLP-forecaster, a K-fold cross-validation was carried out. The results indicate that the proposed model performs well, while the correlation coefficient is in the range 98-99% for sunny days and 94-96% for cloudy days. As an application, the comparison between the forecasted one and the energy produced by the GCPV plant installed on the rooftop of the municipality of Trieste shows the goodness of the proposed model.     

Modeling of solar radiation using remote sensing and artificial neural network in Turkey

Artificial neural networks (ANNs) were used to estimate solar radiation in Turkey (26–45°E, 36–42°N) using geographical and satellite-estimated data. In order to train the Generalized regression neural network (GRNN) geographical and satellite-estimated data for the period from January 2002 to December 2002 from 19 stations spread over Turkey were used in training (ten stations) and testing (nine stations) data. Latitude, longitude, altitude, surface emissivity for ?_4, surface emissivity for ?_5, and land surface temperature are used in the input layer of the network. Solar radiation is the output. Root Mean Square Error (RMSE) and correlation coefficient (R²) between the estimated and measured values for monthly mean daily sum with ANN values have been found as 0.1630 MJ/m² and 95.34% (training stations), 0.3200 MJ/m² and 93.41% (testing stations), respectively. Since these results are good enough it was concluded that the developed GRNN tool can be used to predict the solar radiation in Turkey.     

Modelling of solar energy potential in Nigeria using an artificial neural network model

In this study, an artificial neural network (ANN) based model for prediction of solar energy potential in Nigeria (lat. 4–14°N, log. 2–15°E) was developed. Standard multilayered, feed-forward, back-propagation neural networks with different architecture were designed using neural toolbox for MATLAB. Geographical and meteorological data of 195 cities in Nigeria for period of 10 years (1983–1993) from the NASA geo-satellite database were used for the training and testing the network. Meteorological and geographical data (latitude, longitude, altitude, month, mean sunshine duration, mean temperature, and relative humidity) were used as inputs to the network, while the solar radiation intensity was used as the output of the network. The results show that the correlation coefficients between the ANN predictions and actual mean monthly global solar radiation intensities for training and testing datasets were higher than 90%, thus suggesting a high reliability of the model for evaluation of solar radiation in locations where solar radiation data are not available. The predicted solar radiation values from the model were given in form of monthly maps. The monthly mean solar radiation potential in northern and southern regions ranged from 7.01–5.62 to 5.43–3.54 kW h/m² day, respectively. A graphical user interface (GUI) was developed for the application of the model. The model can be used easily for estimation of solar radiation for preliminary design of solar applications.     

Some suggestions for photovoltaic power generation using artificial light illumination

Part of the electrical energy spent for lighting purposes can be recycled by photovoltaic power generation. We have suggested some methods to enhance power generation based on solar cells using indoor artificial lights. An emergency street light system working on stored photovoltaic energy from outdoor sodium vapor lamp illumination is proposed as an application.     

Electric load analysis using an artificial neural network

Load forecasting in the current, increasingly liberalized, electricity power market is of crucial importance as a means for producers to optimize and rationalize energy supply. A number of electric power companies are equipped to make forecasts with the aid of traditional statistical methods. This paper presents the use of an artificial neural net to an hourly based load forecasting application for a small electric grid on an Italian island (Lipari) not connected to the mainland. The aim was to examine the forecasting ability of a neural net in a situation where the electric load was subject to considerable seasonal variations over the year. The variations are affected by energy demand related to the tourism season as well as by climatic conditions, especially temperature. The network developed was a multi‐layer perceptron type built on three layers trained with a back‐propagation algorithm. The input layer receives all the most relevant information regarding: the class of day type, the hour in the daytime, the load and background temperature recorded at the indicated time for the months of March, August and October whilst the output layer provides the forecast value at the indicated time in December. The results obtained are encouraging; in the training phase the RMS error rate was around 2% and this rate settled at about 2.6% during testing. As both the margins of error recorded are acceptable, the use of a neural network for electric load forecasting applications can be considered an attractive option. Copyright © 2005 John Wiley & Sons, Ltd.     

Performance assessment of electric power generations using an adaptive neural network algorithm

Efficiency frontier analysis has been an important approach of evaluating firms’ performance in private and public sectors. There have been many efficiency frontier analysis methods reported in the literature. However, the assumptions made for each of these methods are restrictive. Each of these methodologies has its strength as well as major limitations. This study proposes a non-parametric efficiency frontier analysis method based on the adaptive neural network technique for measuring efficiency as a complementary tool for the common techniques of the efficiency studies in the previous studies. The proposed computational method is able to find a stochastic frontier based on a set of input–output observational data and do not require explicit assumptions about the function structure of the stochastic frontier. In this algorithm, for calculating the efficiency scores, a similar approach to econometric methods has been used. Moreover, the effect of the return to scale of decision-making units (DMUs) on its efficiency is included and the unit used for the correction is selected by notice of its scale (under constant return to scale assumption). An example using real data is presented for illustrative purposes. In the application to the power generation sector of Iran, we find that the neural network provide more robust results and identifies more efficient units than the conventional methods since better performance patterns are explored. Moreover, principle component analysis (PCA) is used to verify the findings of the proposed algorithm.     

Hour-ahead wind power and speed forecasting using simultaneous perturbation stochastic approximation (SPSA) algorithm and neural network with fuzzy inputs

Wind energy is currently one of the types of renewable energy with a large generation capacity. However, since the operation of wind power generation is challenging due to its intermittent characteristics, forecasting wind power generation efficiently is essential for economic operation. This paper proposes a new method of wind power and speed forecasting using a multi-layer feed-forward neural network (MFNN) to develop forecasting in time-scales that can vary from a few minutes to an hour. Inputs for the MFNN are modeled by fuzzy numbers because the measurement facilities provide maximum, average and minimum values. Then simultaneous perturbation stochastic approximation (SPSA) algorithm is employed to train the MFNN. Real wind power generation and wind speed data measured at a wind farm are used for simulation. Comparative studies between the proposed method and traditional methods are shown.     

SOC estimation method of power battery based on BP artificial neural network

电池剩余电量（SOC）的估算是电池管理系统中的关键技术之一,在众多估算方法中,神经网络在估算的准确性及鲁棒性上具有明显优势。庞大的数据量是获得SOC精确值的重要因素。针对以上问题,研究提出了基于BP人工神经网络的动力电池SOC估算方法,以某型号整包电池作为实验对象,通过对电池电压、电流、内阻及温度的数据采集,获得海量数据。建立电池的等效电路模型,考虑电池极化、充放电倍率及温度的影响对初始数据进行修正。基于MATLAB平台建立BP人工神经网络模型,数据修正后用于网络模型的训练,并验证了模型的可行性。将模型用于实验数据的预测,通过函数拟合实现了SOC的估算。最后,通过对比SOC的预测值与实际测量值,最终证明建立的人工神经网络模型对SOC估算的有效性。     

Forecasting based on neural network approach of solar potential in Turkey

As Turkey lies near the sunny belt between 36 and 42°N latitudes, most of the locations in Turkey receive abundant solar energy. Average annual temperature is 18–20 °C on the south coast, falls down to 14–16 °C on the west coast, and fluctuates 4–18 °C in the central parts. The yearly average solar radiation is 3.6 kW h/m² day, and the total yearly radiation period is 2610 h. The main focus of this study is put forward to solar energy potential in Turkey using artificial neural networks (ANNs). Scaled conjugate gradient (SCG), Pola-Ribiere conjugate gradient (CGP), and Levenberg–Marquardt (LM) learning algorithms and logistic sigmoid transfer function were used in the network. In order to train the neural network, meteorological data for last 4 years (2000–2003) from 12 cities (Çanakkale, Kars, Hakkari, Sakarya, Erzurum, Zonguldak, Balıkesir, Artvin, Çorum, Konya, Siirt, Tekirdağ) spread over Turkey were used as training (nine stations) and testing (three stations) data. Meteorological and geographical data (latitude, longitude, altitude, month, mean sunshine duration, and mean temperature) is used as input to the network. Solar radiation is the output. The maximum mean absolute percentage error was found to be less than 6.78% and R² values to be about 99.7768% for the testing stations. These values were found to be 5.283 and 99.897% for the training stations. The trained and tested ANN models show greater accuracy for evaluating solar resource posibilities in regions where a network of monitoring stations have not been established in Turkey. The predictions from ANN models could enable scientists to locate and design solar energy systems in Turkey and determine the best solar technology.     

Mapping of solar energy potential in Indonesia using artificial neural network and geographical information system

The first objective of this study is to determine the theoretical potential of solar irradiation in Indonesia by using artificial neural networks (ANNs) method. The second objective is to visualize the solar irradiation by province as solar map for the entire of Indonesia. The geographical and meteorological data of 25 locations that were obtained from NASA database are used for training the neural networks and the data from 5 locations were used for testing the estimated values. The testing data were not used in the training of the network in order to give an indication of the performance of the system at unknown locations. In this study, the multi layer perceptron ANNs model, with 9 inputs variables i.e. average temperature, average relative humidity, average sunshine duration, average wind speed, average precipitation, longitude, latitude, latitude, and month of the year were proposed to estimate the monthly solar irradiation as the output. Statistical error analysis in terms of mean absolute percentage error (MAPE) was conducted for testing data to evaluate the performance of ANN model. The best result of MAPE was found to be 3.4% when 9 neurons were set up in the hidden layer. As developing country and wide islands area, Indonesia has the limitation on the number of meteorological station to record the solar irradiation availability; this study shows the ANN method can be an alternative option to estimate solar irradiation data. Monthly solar mapping by province for the entire of Indonesia are developed in GIS environment by putting the location and solar irradiation value in polygon format. Solar irradiation map can provide useful information about the profile of solar energy resource as the input for the solar energy system implementation.     

High power fuel cell simulator based on artificial neural network

Artificial Neural Network (ANN) has become a powerful modeling tool for predicting the performance of complex systems with no well-known variable relationships due to the inherent properties. A commercial Polymeric Electrolyte Membrane fuel cell (PEMFC) stack (5 kW) was modeled successfully using this tool, increasing the number of test into the 7 inputs – 2 outputs-dimensional spaces in the shortest time, acquiring only a small amount of experimental data. Some parameters could not be measured easily on the real system in experimental tests; however, by receiving the data from PEMFC, the ANN could be trained to learn the internal relationships that govern this system, and predict its behavior without any physical equations. Confident accuracy was achieved in this work making possible to import this tool to complex systems and applications.