基于能量熵重构与支持向量回归的水电机组状态趋势预测

针对水电机组运行状态趋势预测的问题,提出了一种基于能量熵重构(EER)与支持向量回归(SVR)的混合预测模型。先针对复杂非平稳监测信号,利用快速集成经验模态分解(FEEMD)算法将其分解为多个本征模态函数(IMFs)分量和单个残余分量;然后基于能量熵(EE)理论对各分量进行重构,以有效降低分量的复杂度;最后,将生成的重构本征模态函数(RIMFs)作为SVR的输入,训练模型参数得到最优的SVR,用于预测机组状态发展趋势。与实例对比分析表明,所提混合预测模型具有较高的预测精度,为机组运维策略的制定提供了一定的指导。     

A deep learning ensemble approach for crude oil price forecasting

As crude oil price is influenced by numerous factors, capturing its behavior precisely is quite challenging, and thus leads to the difficulty of forecasting. In this study, a deep learning ensemble approach is proposed to deal with this problem. In our approach, two techniques are utilized. One is an advanced deep neural network model named stacked denoising autoencoders (SDAE) which is used to model the nonlinear and complex relationships of oil price with its factors. The other is a powerful ensemble method named bootstrap aggregation (bagging) which generates multiple data sets for training a set of base models (SDAEs). Our approach combines the merits of these two techniques and is especially suitable for oil price forecasting. In the empirical study, the WTI crude oil price series are investigated and 198 economic series are used as exogenous variables. Our approach is tested against some competing approaches and shows superior forecasting ability that is statistically proved by three tests.     

城市用水量组合预测模型及其应用

针对城市用水量时间序列包含逐步增长趋势、季节性趋势及不确定性的非线性波动特点,单一预测模型往往很难充分反映原始数据中全部的有效信息,结合季节性时间序列模型(SARIMA)和BP神经网络二者优点,构建了一种新型的组合预测模型,对上海市用水量进行不同时间尺度的预测。结果表明,在不同时间尺度上组合预测模型均比单一预测模型精度高、预测质量稳定。     

An effective and robust decomposition-ensemble energy price forecasting paradigm with local linear prediction

Energy price time series exhibit nonlinear and nonstationary features, which make accurate forecasting energy prices challenging. In this paper, we propose a novel decomposition-ensemble forecasting paradigm based on ensemble empirical mode decomposition (EEMD) and local linear prediction (LLP). The EEMD is used to decompose energy price time series into components, including several intrinsic mode functions and one residual with a simplified structure. Motivated by the findings of the fully local characteristics of a time series decomposed by the EEMD, we adopt the LLP technique to forecast each component. The forecasting results of all the components are aggregated as a final forecast. For validation, three types of energy price time series, crude oil, electricity and natural gas prices, are studied. The experimental results indicate that the proposed model achieves an improvement in terms of both level forecasting and direction forecasting. The performance of the proposed model is also validated through comparison with several energy price forecasting approaches from the literature. In addition, the robustness and the effects of the parameter settings of LLP are investigated. We conclude the proposed model is easy to implement and efficient for energy price forecasting.     

基于EEMD-ACS-SELM的弃风电量组合预测模型

风电已在电力系统中得到了有效利用,因此,弃风电量的准确预测对于电网的安全、经济运行至关重要。文章提出了一种基于集合经验模态分解(EEMD)和t分布自适应变异布谷鸟算法(ACS)优化改进极限学习机(SELM)的弃风电量组合预测方法(EEMD-ACS-SELM)。该方法先采用集合经验模态分解,将原始弃风电量序列分解为一系列不同频率的分量,基于模糊熵理论计算各分量的熵值,并将熵值相似序列重构为新的子序列。然后,将新序列分别建立改进极限学习机预测模型,利用ACS优化算法对SELM算法的输入权值和阈值进行优化。最后,将各序列预测值叠加求和得到原始弃风电量序列的预测值。以新疆某风电场实际运行数据进行算例分析,结果表明,文章所提方法对弃风电量的预测具有较高的精度。     

基于信息融合和堆叠模型的超短期风电功率预测

针对超短期风电功率预测,准确捕捉功率变化因素和建立混合预测模型是提高预测精度的有效手段之一。为了能够继承和整合单个模型的优点以及增强历史信息的表示和利用能力,文章提出了一种基于信息融合和堆叠模型的超短期风电功率预测模型。首先,利用相关性方法选择历史功率序列和历史测风塔数据的特征,作为预测模型的输入;然后,建立两层堆叠的集成模型作为预测模型,并使用交叉验证和超参数优化以增强预测模型的泛化性能;最后,以每个基学习器的输出作为元学习器获得最终预测值的新输入。通过东北某风电场真实数据的验证,以及与单一模型、深度神经网络模型和集成学习模型的对比,验证了所提模型的可行性和有效性。     

基于集合经验模态分解和自回归滑动平均的某碾压混凝土重力坝变形预测模型及应用

为更精确地预测大坝变形数据,针对大坝变形监测序列的非线性和非平稳性问题,提出了一种结合集合经验模态分解和自回归滑动平均模型的大坝变形预测模型。首先利用集合经验模态分解法将非平稳的大坝变形监测数据分解为具有不同特征尺度的本征模态函数,然后分析各分量特征并分别建立自回归滑动平均模型,选择各自适合的最优模型参数,最后叠加各分量的预测结果作为最终的变形预测结果。分析结果表明,相较单一预测模型,结合集合经验模态分解和自回归滑动平均模型的组合预测模型的预测精度更高。     

Using machine learning tools for forecasting natural gas consumption in the province of Istanbul

Commensurate with unprecedented increases in energy demand, a well-constructed forecasting model is vital to managing energy policies effectively by providing energy diversity and energy requirements that adapt to the dynamic structure of the country. In this study, we employ three alternative popular machine learning tools for rigorous projection of natural gas consumption in the province of Istanbul, Turkey's largest natural gas-consuming mega-city. These tools include multiple linear regression (MLR), an artificial neural network approach (ANN) and support vector regression (SVR). The results indicate that the SVR is much superior to ANN technique, providing more reliable and accurate results in terms of lower prediction errors for time series forecasting of natural gas consumption. This study could well serve a useful benchmarking study for many emerging countries due to the data structure, consumption frequency, and consumption behavior of consumers in various time-periods.     

Forecasting risks of natural gas consumption in Slovenia

Efficient operation of modern energy distribution systems often requires forecasting future energy demand. This paper proposes a strategy to estimate forecasting risk. The objective of the proposed method is to improve knowledge about expected forecasting risk and to estimate the expected cash flow in advance, based on the risk model. The strategy combines an energy demand forecasting model, an economic incentive model and a risk model. Basic guidelines are given for the construction of a forecasting model that combines past energy consumption data, weather data and weather forecast. The forecasting model is required to estimate expected forecasting errors that are the basis for forecasting risk estimation. The risk estimation strategy also requires an economic incentive model that describes the influence of forecasting accuracy on the energy distribution systems’ cash flow. The economic model defines the critical forecasting error levels that most strongly influence cash flow. Based on the forecasting model and the economic model, the development of a risk model is proposed. The risk model is associated with critical forecasting error levels in the context of various influential parameters such as seasonal data, month, day of the week and temperature. The risk model is applicable to estimating the daily forecasting risk based on the influential parameters. The proposed approach is illustrated by a case study of a Slovenian natural gas distribution company.     

梯级水电站群超短期入库流量滞时时间序列预报方法

超短期预报调度对于实现短期调度和实时调度无缝衔接、降低调度风险、提高水电调度水平意义重大。而实际超短期预报调度面临着有效预报信息短缺、实时采集信息错报漏报、支流小水电调节、模型通用性差等一系列问题,为此,提出了一种梯级水电站群超短期滞时时间序列预报方法,首先分析了上游电站历史出库和下游电站历史入库流量,利用相关系数找出最强滞时流量匹配关系,从而将上游电站出库流量从下游电站入库中分解出来,还原出有效的区间流量,再通过区间流量逐日化进行时间序列建模,最后与上游电站滞时出库合成为下游电站预报入库流量。以云南澜沧江干流梯级电站为例,取得了良好的预测效果,从而验证了该方法的合理性、准确性与通用性。     

基于小波分析和灰色模型的用电量预测

月度用电量同时具有增长性和季节波动性的二重趋势,针对月度用电量的这一变化特点,提出了一种基于小波分析和灰色预测模型的用电量预测方法,同时考虑春节影响因素,结合移位修正法对1月份和2月份的用电量进行修正.经过实例分析和计算,结果表明该方法有较高的预测精度,具有较好的适用性和可行性.     

A seasonal hybrid procedure for electricity demand forecasting in China

Electricity is a special energy which is hard to store, so the electricity demand forecasting in China remains an important problem. This paper aims at developing an improved hybrid model for electricity demand in China, which takes the advantages of moving average procedure, combined method, hybrid model and adaptive particle swarm optimization algorithm, known as MA-C-WH. It is designed for making trend and seasonal adjustments which simultaneously presents the electricity demand forecasts. Four actual electricity demand time series in China power grids are selected to illustrate the proposed MA-C-WH model, and one existing seasonal autoregressive integrated moving average model (SARIMA) is selected to compare with the proposed model using the same data series. The results of popular forecasting precision indexes show that our proposed model is an effective forecasting technique for seasonal time series with nonlinear trend.     

基于改进LSSVM的太阳辐射量预测

为准确预测太阳辐射量,提出一种基于变分模态分解和粒子群优化算法的最小二乘支持向量机组合预测模型。针对太阳辐射量序列具有不稳定性的特点,首先利用变分模态分解将历史太阳辐射量数据分解成一系列相对稳定的分量序列,再应用粒子群优化最小二乘支持向量机参数,以预测各分量序列,将各分量太阳辐射量预测值集成,从而得到最终太阳辐射量预测值。实例分析和对比研究表明,该模型预测太阳辐射量有效可行,具有较高的预测精度。研究成果可为太阳辐射量预测提供参考。     

基于Kmeans和CEEMD-PE-LSTM的短期光伏发电功率预测

针对光伏发电功率时间序列随机性和波动性强的特点,提出一种基于Kmeans和完备总体经验模态分解(CEEMD)、排列熵(PE)、长短期记忆(LSTM)神经网络结合的短期光伏功率预测模型。先通过Kmeans算法选出预测日的相似日;然后采用CEEMD将发电功率和影响因素数据的原始序列分解为多个固有模态分量,并用排列熵算法对模态分量进行重构;最后对重构后的子序列分别进行LSTM建模预测,再将子序列预测结果叠加起来确定光伏发电功率预测值。试验结果表明,所提预测模型与单独的LSTM预测模型和EMD-PE-LSTM预测模型相比,功率预测精度明显提高,为电网调度提供了一定参考。     

基于分数阶灰色Elman组合模型的中长期负荷预测

针对电力系统中长期负荷预测样本少、间隔时间长、影响因素多等问题,提出基于分数阶灰色Elman的组合预测模型,首先针对负荷预测样本少、增长趋势明显的特点,利用分数阶灰色模型弱化原始序列的随机性,降低解的扰动界,其次利用Elman神经网络模型适应性与学习能力强的特点来解决负荷的非线性及影响因素复杂的问题,然后根据最优模型赋予二者最优权值,得到最终的组合模型,最后采用傅里叶级数残差校正模型修正组合模型的误差。仿真结果表明,本文提出的方法可有效拟合负荷的变化趋势,提升负荷预测的准确度。     

Time series models (Grey-Markov,Grey Model with rolling mechanism and singular spectrum analysis) to forecast energy consumption in India

The present study applies three time series models, namely, Grey-Markov model, Grey-Model with rolling mechanism, and singular spectrum analysis (SSA) to forecast the consumption of conventional energy in India. Grey-Markov model has been employed to forecast crude-petroleum consumption while Grey-Model with rolling mechanism to forecast coal, electricity (in utilities) consumption and SSA to predict natural gas consumption. The models for each time series has been selected by carefully examining the structure of the individual time series. The mean absolute percentage errors (MAPE) for two out of sample forecasts have been obtained as follows: 1.6% for crude-petroleum, 3.5% for coal, 3.4% for electricity and 3.4% for natural gas consumption. For two out of sample forecasts, the prediction accuracy for coal consumption was 97.9%, 95.4% while for electricity consumption the prediction accuracy was 96.9%, 95.1%. Similarly, the prediction accuracy for crude-petroleum consumption was found to be 99.2%, 97.6% while for natural gas consumption these values were 98.6%, 94.5%. The results obtained have also been compared with those of Planning Commission of India's projection. The comparison clearly points to the enormous potential that these time series models possess in energy consumption forecasting and can be considered as a viable alternative.     

基于小波分析的BP网络预测模型及其在年径流预测中的应用

鉴于小波变换序列中尺度系数系列和小波系数系列变化特征存在较大差异,提出了一种新的小波分析与BP网络结合方式,即建立两个BP网络分别对两类系数系列进行预测,再对各小波变换系数的预测值进行小波重构,获得原序列的预测值。将该模型应用于二滩电站入库年径流量预测,结果表明该模型预测精度高,可为水电站提供可靠的入库年径流预测结果。     

基于STL Former的中短期光伏功率预测

由于光伏发电具有间歇性和波动性,给电网运行的安全性和稳定性造成危害,对光伏功率进行准确预测可以有效解决这一问题。本文提出一种基于STL Former的中短期光伏功率预测模型,该模型结合了季节趋势局部加权回归分解（STL分解）与神经网络模型。首先,STL Former模型将光伏功率数据通过STL分解进行特征扩充,用于提取基于历史序列的周期项、趋势项特征。然后,拼接周期项、趋势项特征和原特征,进行数据预处理和特征编码并使用基于Informer模型的神经网络进行功率预测。最后,在真实数据集上进行大量实验。实验结果表明：STL Former在中短期光伏功率预测任务中精度较高,其中在2 h光伏功率预测任务时,平均绝对值误差为0.176、均方误差为0.180;在28 h光伏功率预测任务时,平均绝对值误差为0.170、均方误差为0.154。     

On the importance of the long-term seasonal component in day-ahead electricity price forecasting

In day-ahead electricity price forecasting (EPF) the daily and weekly seasonalities are always taken into account, but the long-term seasonal component (LTSC) is believed to add unnecessary complexity to the already parameter-rich models and is generally ignored. Conducting an extensive empirical study involving state-of-the-art time series models we show that (i) decomposing a series of electricity prices into a LTSC and a stochastic component, (ii) modeling them independently and (iii) combining their forecasts can bring – contrary to a common belief – an accuracy gain compared to an approach in which a given time series model is calibrated to the prices themselves.     

A simulated-based neural network algorithm for forecasting electrical energy consumption in Iran

This study presents an integrated algorithm for forecasting monthly electrical energy consumption based on artificial neural network (ANN), computer simulation and design of experiments using stochastic procedures. First, an ANN approach is illustrated based on supervised multi-layer perceptron (MLP) network for the electrical consumption forecasting. The chosen model, therefore, can be compared to that of estimated by time series model. Computer simulation is developed to generate random variables for monthly electricity consumption. This is achieved to foresee the effects of probabilistic distribution on monthly electricity consumption. The simulated-based ANN model is then developed. Therefore, there are four treatments to be considered in analysis of variance (ANOVA), which are actual data, time series, ANN and simulated-based ANN. Furthermore, ANOVA is used to test the null hypothesis of the above four alternatives being statistically equal. If the null hypothesis is accepted, then the lowest mean absolute percentage error (MAPE) value is used to select the best model, otherwise the Duncan method (DMRT) of paired comparison is used to select the optimum model which could be time series, ANN or simulated-based ANN. In case of ties the lowest MAPE value is considered as the benchmark. The integrated algorithm has several unique features. First, it is flexible and identifies the best model based on the results of ANOVA and MAPE, whereas previous studies consider the best fitted ANN model based on MAPE or relative error results. Second, the proposed algorithm may identify conventional time series as the best model for future electricity consumption forecasting because of its dynamic structure, whereas previous studies assume that ANN always provide the best solutions and estimation. To show the applicability and superiority of the proposed algorithm, the monthly electricity consumption in Iran from March 1994 to February 2005 (131 months) is used and applied to the proposed algorithm.