Multi-step-ahead forecasting of daily solar radiation components in the Saharan climate

ABSTRACT

Accurate estimation of renewable energy sources plays an important role in their integration into the grid. An unexpected atmospheric change can produce a range of problems related to various solar plant components affecting the electricity generation system. Global solar radiation (GSR) assessment has been increased in the past decade due to its important use in photovoltaic application. In this paper, we propose the use of machine learning-based models for daily global and direct solar radiation forecasting in a semi-arid climate, using a combination set of meteorological parameters on a horizontal surface in the Ghardaïa region. The models are presented and implemented on 3-year measured meteorological data at Applied Research Unit for Renewable Energies (URAER) at Ghardaïa city between 2014 and 2016. The results show that both MLP and RBF models perform well for three-step-ahead forecasting with a slight improvement in MLP models in terms of statistical metrics.     

太阳散射辐射逐日预测模型研究

针对于传统的确定性太阳辐射模型不能反映气象变化的弊端,提出了基于回归BP神经网络和小波分析理论的太阳散射辐射逐日预测模型。神经网络具有非线性函数逼近及自组织自学习的能力,基于小波分析在信号处理方面的时频域多分辨特性,本文利用小波变换将太阳散射辐射数据序列进行时频域分解后作为神经网络预测模型的输入样本,实例表明该方法与传统模型相比预测精度高,具有可行性。     

Decomposing global solar radiation into its diffuse and direct normal radiation

ABSTRACT

This work presents a model based on Radial Basis Function (RBF) to estimate the diffused solar radiation (DSR) and direct normal radiation (DNR) fractions of solar radiation from global solar radiation in a semiarid area in Algeria based on a database measured between 2013 and 2015. The data has been collected at Applied Research Unit for Renewable Energies, (URAER) at Ghardaia city situated in the south of Algeria. The experimental results show that RBF model estimates DNR and DSR with high performance. The difference between the measured and the predicted values show a normalised Root Mean Square Error (nRMSE) of 0.033 and 0.065 for DNR and DSR, respectively. The obtained values of Determination Coefficient (R²) and Correlation Coefficient (R) are: 97.3%, 98.60%, respectively for DNR and 88.89%, 91.12% For DSR.

However, the obtained results are very plausible and showed that RBF model estimates the DSR and DNR with good accuracy.     

Hybrid ensemble soft computing approach for predicting penetration rate of tunnel boring machine in a rock environment

This study implements a hybrid ensemble machine learning method for forecasting the rate of penetration (ROP) of tunnel boring machine (TBM), which is becoming a prerequisite for reliable cost assessment and project scheduling in tunnelling and underground projects in a rock environment. For this purpose, a sum of 185 datasets was collected from the literature and used to predict the ROP of TBM. Initially, the main dataset was utilised to construct and validate four conventional soft computing (CSC) models, i.e. minimax probability machine regression, relevance vector machine, extreme learning machine, and functional network. Consequently, the estimated outputs of CSC models were united and trained using an artificial neural network (ANN) to construct a hybrid ensemble model (HENSM). The outcomes of the proposed HENSM are superior to other CSC models employed in this study. Based on the experimental results (training RMSE = 0.0283 and testing RMSE = 0.0418), the newly proposed HENSM is potential to assist engineers in predicting ROP of TBM in the design phase of tunnelling and underground projects.     

Controlling power output of solar chimney power plant according to demand

ABSTRACT

The solar chimney power plant (SCPP) is a simple solar thermal power plant that is capable of converting solar energy into thermal energy in the solar collector. In the second stage, the generated thermal energy is converted into kinetic energy in the chimney and ultimately into electric energy using a combination of a wind turbine and a generator. The numerical simulations were performed for the geometry of the prototype in Manzanares, Spain. Using computational ?uid dynamics (CFD) techniques; we have simulated a two-dimensional axisymmetric model of a SCPP with the RNG k-ε turbulence. In this model, the discrete ordinates (DO) radiation model was implemented to solve the radiative transfer equation, using a two-band radiation model. The main objective of this work is to explore dynamic control over plant power output. We have presented a technique to control the power output of the solar chimney power plant, in order to deliver power according to specified demand patterns. In order to present this, the reference plant model was modified to include a secondary and tertiary collector roof under the existing main collector. In terms of base load electricity generation, the inclusion of a secondary and tertiary collector roof produces good control over plant output.     

基于LS-SVM的粉煤灰混凝土强度预测模型

提出一种基于最小二乘支持向量机(LS-SVM)的粉煤灰混凝土强度智能预测模型,并给出了相应的步骤和算法。通过该模型分析了水胶比、水泥用量、粉煤灰替代率及砂率等因素对粉煤灰混凝土强度的影响。在此基础上,对不同配比所浇注的混凝土强度进行预测,有助于准确认识混凝土强度随配比参数的变化规律。与多元线性回归、神经网络及标准SVM模型比较,该模型的优点为:(1)采用了结构风险最小化准则,在最小化样本误差的同时减小模型泛化误差的上界,提高了模型小样本泛化能力;(2)将迭代学习算法转换为求解线性方程组,使得整个模型仅有一个全局最优点,解决局部最小问题;(3)用等式约束代替标准SVM算法中的不等式约束,将求解二次规划问题转化为直接求解线性矩阵方程,有效提高建模速度。用该模型对混凝土的强度预测实例表明,其建模速度比标准SVM高近1个数量级,预测误差仅为SVM方法的20%、BP神经网络方法的10%左右。     

Intercomparison of air quality data using principal component analysis, and forecasting of PM₁₀ and PM₂.₅ concentrations using artificial neural networks, in Thessaloniki and Helsinki

In this paper we propose a methodology consisting of specific computational intelligence methods, i.e. principal component analysis and artificial neural networks, in order to inter-compare air quality and meteorological data, and to forecast the concentration levels for environmental parameters of interest (air pollutants). We demonstrate these methods to data monitored in the urban areas of Thessaloniki and Helsinki in Greece and Finland, respectively. For this purpose, we applied the principal component analysis method in order to inter-compare the patterns of air pollution in the two selected cities. Then, we proceeded with the development of air quality forecasting models for both studied areas. On this basis, we formulated and employed a novel hybrid scheme in the selection process of input variables for the forecasting models, involving a combination of linear regression and artificial neural networks (multi-layer perceptron) models. The latter ones were used for the forecasting of the daily mean concentrations of PM₁₀ and PM_2.5 for the next day. Results demonstrated an index of agreement between measured and modelled daily averaged PM₁₀ concentrations, between 0.80 and 0.85, while the kappa index for the forecasting of the daily averaged PM₁₀ concentrations reached 60% for both cities. Compared with previous corresponding studies, these statistical parameters indicate an improved performance of air quality parameters forecasting. It was also found that the performance of the models for the forecasting of the daily mean concentrations of PM₁₀ was not substantially different for both cities, despite the major differences of the two urban environments under consideration.     

Development of the typical meteorological database for Chinese locations

The typical meteorological database for 57 Chinese locations was developed for building simulations and air-conditioning design. The database consists of three parts: the typical meteorological years (TMY), the typical meteorological days (TMD) and the design temperature and humidity (DTH). The typical meteorological year (TMY) is the main part of the database. Because there are not data on solar radiation in the observations, a method to estimate solar radiation with dry bulb temperature difference, relative humidity, total cloud cover and wind speed was developed. Methodologies of interpolations were developed to produce 1 h data with the 3 h data. The global solar radiation on the horizontal surface was separated into direct and diffuse components with the Gompertz function. The typical meteorological day (TMD) consists of the monthly average values of dry bulb temperature, solar radiation, relative humidity, etc. for each hour of the day. The design temperature and humidity (DTH) was developed for the purpose of air-conditioning design. The frequencies of 2.5% and 5.0% were selected to decide the design temperature and humidity for the 57 Chinese locations.     

Generation of typical weather data using the ISO Test Reference Year (TRY) method for major cities of South Korea

Computer simulation of buildings and solar energy systems is being used increasingly in energy assessments and design. Simulation often requires hourly weather data. Such data sets are the Test Reference Years (TRYs), Typical Meteorological Year (TMY) and Weather Year for Energy Calculations (WYEC). Typical weather data consists of 8760 values of various selected meteorological parameters such as ambient temperature, solar radiation, relative humidity and wind velocity and are originally derived from long-term data. This paper discusses methods of selecting typical weather data, the possibility of using the cloud cover data instead of daily global radiation and describes the selection of ISO Test Reference Year (TRY) for major cities of South Korea. The ISO-15927 procedure and algorithms are explained in detail and the Finkelstein–Schafer statistic, the basic selector statistic explained. ISO TRYs for the major cities of South Korea are derived from 20 years of meteorological data recorded during the period 1986–2005. A comparison is made between the 7 sites demonstrating the link between dry-bulb temperature, solar radiation and latitude.     

典型气象年和典型代表年的选择及其对建筑能耗的影响

介绍了典型气象年和典型代表年的选择原理和几种常见的选择方法。不同的方法考虑了不同气象参数的加权因子和气象数据的连续性。介绍了将太阳辐射总量分为太阳直射辐射量与太阳散射辐射量的应用模型,并依据香港的气象数据,分别计算选出了香港的典型气象年与典型代表年。为了验证不同方法计算出的典型气象年与典型代表年对研究对象、系统的影响,作了一个实例建筑物能耗动态模拟。结果表明,不同典型气象年对模拟结果的影响偏差较小,而典型代表年的影响较大;选择合适方法计算的典型气象年对保证模拟评估结果的正确性具有重要意义。     

Development of solar heat gain factors database using meteorological data

In tropical and subtropical regions, solar heat gain via fenestration, particularly on vertical surfaces, plays an important role in determining the thermal performance of a building. For sizing air-conditioning equipment, maximum solar heat gain factors (SHGFs) are used for estimating the design peak load. Recently, the SHGF data representing the prevailing weather conditions have become essential and more practical for part load performance designs and daylighting schemes evaluation. In the absence of measured solar radiation data for the determination of SHGFs, meteorological radiation models may be used. This paper presents the validation of SHGFs prediction models based on sunshine hours and horizontal solar data. Statistical assessments for the models have shown that using sunshine hours to predict the hourly SHGFs may not be appropriate for dynamic building simulation studies. For the average SHGFs computation, all models present acceptable results. In determining the SHGFs for horizontal and vertical surfaces at the peak and other significant levels, all prediction models perform better than the ASHRAE clear sky approach, particularly at high significant levels. This finding also provides information for the estimation of total air-conditioning plant capacity at both the peak load operation and the multiple equipment sizing under part load condition.     

建筑节能分析太阳总辐射模型研究综述

太阳辐射是建筑节能分析的重要基础气象参数,实测数据远远不能满足需求,理论计算是目前获取辐射数据的主要途径。将常用水平面太阳总辐射模型归纳为气象参数、空间插值和基于DEM三类,详述了各自的原理和计算方法。对三类模型在建筑节能分析中的适用性进行了分析,展望了建筑节能分析用太阳辐射模型的发展趋势:气象参数模型与DEM模型的融合。     

Forecasting short-term electricity load using hybrid support vector regression with grey catastrophe and random forest modeling

This paper develops a novel short-term load forecasting model that hybridizes several machine learning methods, such as support vector regression (SVR), grey catastrophe (GC (1,1)), and random forest (RF) modeling. The modeling process is based on the minimization of both SVR and risk. GC is used to process and extract catastrophe points in the long term to reduce randomness. RF is used to optimize forecasting performance by exploiting its superior optimization capability. The proposed SVR-GC-RF model has higher forecasting accuracy (MAPE values are 6.35% and 6.21%, respectively) using electric loads from Australian-Energy-Market-Operator; it can provide analytical support to forecast electricity consumption accurately.     

Using extreme learning machines for short-term urban water demand forecasting

This study explores the ability of various machine learning methods to improve the accuracy of urban water demand forecasting for the city of Montreal (Canada). Artificial Neural Network (ANN), Support Vector Regression (SVR) and Extreme Learning Machine (ELM) models, in addition to a traditional model (Multiple linear regression, MLR) were developed to forecast urban water demand at lead times of 1 and 3 days. The use of models based on ELM in water demand forecasting has not previously been explored in much detail. Models were based on different combinations of the main input variables (e.g., daily maximum temperature, daily total precipitation and daily water demand), for which data were available for Montreal, Canada between 1999 and 2010. Based on the squared coefficient of determination, the root mean square error and an examination of the residuals, ELM models provided greater accuracy than MLR, ANN or SVR models in forecasting Montreal urban water demand for 1 day and 3 days ahead, and can be considered a promising method for short-term urban water demand forecasting.     

The impact of hourly solar radiation model on building energy analysis in different climatic regions of Turkey

The purpose of this study is to investigate the effect of solar radiation models on the determination of energy performance of a single-family house assisted with renewable energy system including photovoltaic panels and solar water heater. An Angström-Prescott type solar radiation model was compared with Zhang and Huang model derived based on hourly meteorological data of 12 locations in Turkey. Since regression coefficients of the Zhang and Huang model are valid for China, new regression coefficients were derived by using local meteorological data. A clear distinction could not be observed in simulated annual heating load intensity for each model since the average relative deviation of the models’ results was 2.5%. However, the average deviation was 12.5% for space cooling load intensity. Primary energy ratings (PER) and the renewable energy ratio (RER) were determined for each location. For total PER, the highest deviation was 4.6% and 3.3% for Mersin and Mu?la, respectively. For the other locations, this parameter deviates between 0.02%–2.11%. The highest RER was 18.6% for Mersin.     

Direct solar radiation data as input into mathematical models describing the thermal performance of buildings—II. Development of relationships

Solar radiation is not commonly recorded in the meteorological network and it is therefore generally necessary to predict its value theoretically. The majority of solar radiation predictive methods assume a standard atmosphere. Thus they are of limited application to ‘all-year’ thermal design models. A review of the nature of direct solar radiation and methods by which it can be predicted are presented.     

New correlations for estimation of monthly average daily solar radiation on a horizontal surface using meteorological data

This study deals with the estimation of monthly average daily global solar radiation incident on a horizontal surface at a location using meteorological data for different cities of Andhra Pradesh state. The regression equations of two types of models are developed for various locations of the state having varied climatic regions using measured meteorological parameters obtained from different measuring stations. Parameters such as the latitude and height of the location, maximum and minimum temperature and relative humidity are considered to develop linear and quadratic regression equations to estimate the monthly average daily global radiation. The estimated values are compared with measured data and with other theoretical models in terms of mean percent errors calculated using recognised statistical criteria of MBE, MABE, MPE and RMSE. The proposed quadratic model performs better than the proposed linear models and shows good agreement with the long-term pyranometer data of various locations with a deviation of less than 10%. In comparison with other theoretical models presented so far, the suggested models are expected to perform with a higher degree of accuracy.     

An ARIMA-NARX hybrid model for forecasting urban water consumption (case study: Tehran metropolis)

ABSTRACT

In this research, an ARIMA-NARX (Autoregressive Integrated Moving Average-Nonlinear Auto-Regressive eXogenous) hybrid model is proposed to forecast daily Urban Water Consumption (UWC) for Tehran Metropolis. The linear and nonlinear component of the UWC was forecast by ARIMA as a linear forecasting model and the artificial neural network as a nonlinear forecasting model, respectively. An alternative hybrid model including sunshine hour in addition to the previous studies’ predictors (the minimum, maximum and average temperature, relative humidity and precipitation) was selected as the superior alternative model. Then, the performance of proposed model was compared with ARIMA and NARX models. The results showed that the hybrid model, which benefits from capability of both linear and nonlinear models, has a higher accuracy than the other two models in forecasting UWC. Therefore, the proposed hybrid model has better results in UWC forecasting and, as a consequence, better urban water reservoir management will be provided.     

滑坡监控信息分析中的修正灰色系统预测模型及应用

在灰色系统全数据GM(1,1)模型基础上,经过新陈代谢过程、分阶残差修正和权重修正,依次建立了多个不同修正方式的模型,建模过程考虑了实测信息的非等间隔性。结合富阳来龙山滑坡监控信息,对位于滑坡体主滑段,历时4个典型滑移阶段的B14测孔滑带土位移作预测分析,经过各模型精度检验比较表明,所提出的二阶残差权重修正GM(1,1)模型大幅提高了预测精度和适应性。     

Prediction of flyrock induced by mine blasting using a novel kernel-based extreme learning machine

Blasting is a common method of breaking rock in surface mines. Although the fragmentation with proper size is the main purpose, other undesirable effects such as flyrock are inevitable. This study is carried out to evaluate the capability of a novel kernel-based extreme learning machine algorithm, called kernel extreme learning machine (KELM), by which the flyrock distance (FRD) is predicted. Furthermore, the other three data-driven models including local weighted linear regression (LWLR), response surface methodology (RSM) and boosted regression tree (BRT) are also developed to validate the main model. A database gathered from three quarry sites in Malaysia is employed to construct the proposed models using 73 sets of spacing, burden, stemming length and powder factor data as inputs and FRD as target. Afterwards, the validity of the models is evaluated by comparing the corresponding values of some statistical metrics and validation tools. Finally, the results verify that the proposed KELM model on account of highest correlation coefficient (R) and lowest root mean square error (RMSE) is more computationally efficient, leading to better predictive capability compared to LWLR, RSM and BRT models for all data sets.