Fuzzy interaction regression for short term load forecasting

Electric load forecasting is a fundamental business process and well-established analytical problem in the utility industry. Due to various characteristics of electricity demand series and the business needs, electric load forecasting is a classical textbook example and popular application field in the forecasting community. During the past 30 plus years, many statistical and artificial intelligence techniques have been applied to short term load forecasting (STLF) with varying degrees of success. Although fuzzy regression has been tried for STLF for about a decade, most research work is still focused at the theoretical level, leaving little value for practical applications. A primary reason is that inadequate attention has been paid to the improvement of the underlying linear model. This application-oriented paper proposes a fuzzy interaction regression approach to STLF. Through comparisons to three models (two fuzzy regression models and one multiple linear regression model) without interaction effects, the proposed approach shows superior performance over its counterparts. This paper also offers critical comments to a notable but questionable paper in this field. Finally, tips for practicing forecasting using fuzzy regression are discussed.     

A hybrid metaheuritic technique developed for hourly load forecasting

Electricity load forecasting has become one of the most functioning tools in energy efficiency and load management and utility companies which has been made very complex due to deregulation. Due to the importance of providing a secure and economic electricty for the consumers, having a reliable and robust enough forecast engine in short‐term load management is very needful. Fuzzy inference system is one of primal branches of Artificial Intelligence techniques which has been widely used for different applications of decision making in complex systems. This paper aims to develop a Fuzzy inference system as a main forecast engine for Short term Load Forecasting (STLF) of a city in Iran. However, the optimization of this platform for this special case remains a basic problem. Hence, to address this issue, the Radial Movement Optimization (RMO) technique is proposed to optimize the whole Fuzzy platform. To support this idea, the accuracy of the proposed model is analyzed using MAPE index and an average error of 1.38% is obtained for the forecast load demand which represents the reliability of the proposed method. Finally, results achieved by this method, demonstrate that an adaptive two‐stage hybrid system consisting of Fuzzy & RMO can be an accurate and robust enough choice for STLF problems. © 2016 Wiley Periodicals, Inc. Complexity 21: 521–532, 2016     

基于熵值法的组合预测模型及其在高峰负荷预测中应用

组合预测可以综合利用各单一预测方法所提供的信息,是提高预测精度的有效途径.本文在指数平滑预测法及灰色预测方法的基础上建立组合预测模型,采用熵值法确定组合权系数,并对某电网高峰负荷进行了预测.实例表明,此模型具有很强的实用性和很高的预测精度.     

Electric load forecasting by support vector model

Accurately electric load forecasting has become the most important management goal, however, electric load often presents nonlinear data patterns. Therefore, a rigid forecasting approach with strong general nonlinear mapping capabilities is essential. Support vector regression (SVR) applies the structural risk minimization principle to minimize an upper bound of the generalization errors, rather than minimizing the training errors which are used by ANNs. The purpose of this paper is to present a SVR model with immune algorithm (IA) to forecast the electric loads, IA is applied to the parameter determine of SVR model. The empirical results indicate that the SVR model with IA (SVRIA) results in better forecasting performance than the other methods, namely SVMG, regression model, and ANN model.     

基于修正的时间序列模型对冲击负荷的预测

电力负荷预测的实质是对电力市场需求的预测,是利用以往的历史数据资料找出电力负荷的变化规律,进而预测负荷在未来时期的变化趋势.由于经济、气候以及工业生产等诸多因素的约束和限制,电力负荷预测精度很难提高.一个好的实用的电力负荷预测模型则要求既能充分利用负荷的历史数据,又能灵活方便地综合考虑其他多种相关因素的影响.提出了回归与自回归模型相结合的时间序列混合回归预测模型,它的待估参数由BP神经网络进行修正,经实例验证,预测效果良好.     

基于Chebyshev多项式的神经网络中长期负荷预测研究

高精度负荷预测在提高电力系统的安全性和经济性方面有着极其重要的意义,而现有的负荷预测方法因参数有限,难以完全反映其内在规律,因而导致预测结果不够准确.为此提出了一种基于Chebyshev多项式神经网络模型的预测方法.该方法使用递推最小二乘法训练神经网络权值系数,以获得高精度的参数估计,从而实现Chebyshev多项式神经网络模型对负荷量的最优拟合,再利用训练好的Chebyshev多项式神经网络模型实现中长期负荷预测.研究结果表明,该方法能较好模拟负荷变化规律,有效提高了负荷预测精度,在电力系统负荷预测中有较大的应用价值.     

基于改进BP神经网络的电力负荷预测

在现有文献研究的基础上,对BP神经网络进行了深入研究,提出了一种新的LAFBP模型,给出了模型的标准BP算法、改进BP算法、权值和阈值的初始化方法.在此基础上,用新的LAFBP模型与传统的标准BP模型对黑龙江省巴彦县的电力负荷进行了预测.预测结果表明,新的LAFBP模型不仅克服了传统的BP模型外推效果不好的缺点,而且在模型的拟合精度、学习时间和学习次数方面明显优于传统的BP模型.     

含有压缩因子的粒子群优化灰色模型在智能电网中的应用

针对智能电网对用电量预测的需求和电力系统的负荷特性,在分析了灰色模型GM(1,1)的局限性以及基本粒子群算法在优化GM(1,1)背景值时所出现的不足的基础上,构建了具有压缩因子K的粒子群算法,以此来改进灰色模型的背景值,提出了含有压缩因子的粒子群优化灰色模型KPSO-GM,并把它用于智能电网中用电量预测。实例证明,该算法具有较高的预测精度,有利于提高智能电网的质量。     

Rainfall forecasting by technological machine learning models

Accurate forecasting of rainfall has been one of the most important issues in hydrological research. Due to rainfall forecasting involves a rather complex nonlinear data pattern; there are lots of novel forecasting approaches to improve the forecasting accuracy. Recurrent artificial neural networks (RNNS) have played a crucial role in forecasting rainfall data. Meanwhile, support vector machines (SVMs) have been successfully employed to solve nonlinear regression and time series problems. This investigation elucidates the feasibility of hybrid model of RNNs and SVMs, namely RSVR, to forecast rainfall depth values. Moreover, chaotic particle swarm optimization algorithm (CPSO) is employed to choose the parameters of a SVR model. Subsequently, example of rainfall values during typhoon periods from Northern Taiwan is used to illustrate the proposed RSVRCPSO model. The empirical results reveal that the proposed model yields well forecasting performance, RSVRCPSO model provides a promising alternative for forecasting rainfall values.     

Forecasting seasonal time series data: a Bayesian model averaging approach

A flexible Bayesian periodic autoregressive model is used for the prediction of quarterly and monthly time series data. As the unknown autoregressive lag order, the occurrence of structural breaks and their respective break dates are common sources of uncertainty these are treated as random quantities within the Bayesian framework. Since no analytical expressions for the corresponding marginal posterior predictive distributions exist a Markov Chain Monte Carlo approach based on data augmentation is proposed. Its performance is demonstrated in Monte Carlo experiments. Instead of resorting to a model selection approach by choosing a particular candidate model for prediction, a forecasting approach based on Bayesian model averaging is used in order to account for model uncertainty and to improve forecasting accuracy. For model diagnosis a Bayesian sign test is introduced to compare the predictive accuracy of different forecasting models in terms of statistical significance. In an empirical application, using monthly unemployment rates of Germany, the performance of the model averaging prediction approach is compared to those of model selected Bayesian and classical (non)periodic time series models.     

基于非等时距加权灰色模型与神经网络的组合预测算法

非等时距预测算法在不等时间间隔序列的趋势分析与预测方面具有重要作用．在传统灰色预测理论的基础上,提出一种基于非等时距加权灰色模型和神经网络的组合预测算法．通过构建非等时距加权灰色预测模型,将原始数据序列的平均值作为累加序列初值,将连续累积函数的积分面积作为背景值,对累加序列进行加权处理,以真实反映时间序列发展对预测结果的影响．在此基础上,引入BP神经网络对灰色预测的残差序列进行修正,进一步提高了预测精度．经算例验证,该算法预测精度达到1级,且高于类似算法．     

A short and mean-term automatic forecasting system––application to textile logistics

In order to reduce their stocks and to limit stock out, textile companies require specific and accurate sale forecasting systems. More especially, textile distribution involves different forecast lead times: mean-term (one year) and short-term (one week in average). This paper presents two new complementary forecasting models, appropriate to textile market requirements. The first model (AHFCCX) allows to automatically obtain mean-term forecasting by using fuzzy techniques to quantify influence of explanatory variables. The second one (SAMANFIS), based on a neuro-fuzzy method, performs short-term forecasting by readjusting mean-term model forecasts from load real sales. To evaluate forecasts accuracy, our models and classical ones are compared to 322 real items sales series of an important ready to wear distributor.     

电力负荷预测的遗传神经网络模型研究

根据电力负荷预测的特点,提出遗传神经网络负荷预测模型,有效地克服了人工神经网络学习速度慢、存在局部极小点的固有缺陷,经实例验证,该方法能有效地提高预测精度和速度。     

模糊处理变结构神经网络日负荷预测方法研究

对于受不确定因素影响的日电力负荷,首次提出了基于模糊分类规则的变结构神经网络负荷预测模型,考虑从两方面改进预测精度,一个方面是通过模糊分类规则,使过去的负荷数据分为不同气候特征,选用同类特征数据进行预测,另一方面是通过神经网络变结构优化,确定最优网络和最优拟合逼近,从而得到最优的预测结果,这种新方法同时考虑了天气因素的影响和神经网络的最优确定,因此,较大提高了日负荷预测的精度。     

含多抽水蓄能电站的电网多目标运行优化研究

广州抽水蓄能电站和惠州抽水蓄能电站开停机计划不够合理,导致两个蓄能电站不能实现均衡使用。针对此问题,提出在参照预测的日负荷曲线的基础上,利用加速变步长算法确定两个电站的调度方案,并运用回归分析预测法对每个电站的日抽水量进行预测。据此,可以合理安排两个蓄能电站的工作位置,避免机组的频繁启停,使其能够更好地发挥调峰填谷、调频、调压、提供备用等作用。     

Bayesian forecasting with the Holt–Winters model

Exponential smoothing methods are widely used as forecasting techniques in inventory systems and business planning, where reliable prediction intervals are also required for a large number of series. This paper describes a Bayesian forecasting approach based on the Holt–Winters model, which allows obtaining accurate prediction intervals. We show how to build them incorporating the uncertainty due to the smoothing unknowns using a linear heteroscedastic model. That linear formulation simplifies obtaining the posterior distribution on the unknowns; a random sample from such posterior, which is not analytical, is provided using an acceptance sampling procedure and a Monte Carlo approach gives the predictive distributions. On the basis of this scheme, point-wise forecasts and prediction intervals are obtained. The accuracy of the proposed Bayesian forecasting approach for building prediction intervals is tested using the 3003 time series from the M3-competition.     

基于LM算法的BP神经网络的电力负荷短期预测

通过对BP神经网络输入负荷值的归一化处理,同时采用Levenberg-Marquardt(LM)算法,建立了一个改进了的BP神经网络,同时用它来对电力系统进行短期负荷预测.LM算法有效地提高了BP神经网络的收敛速度和负荷的预测精度.仿真结果表明,改进了的BP神经网络具有很高的预测精度和较强的适用能力.     

A second-order accurate three sub-step composite algorithm for structural dynamics

In this paper, a novel three sub-step composite algorithm with desired numerical properties is developed. The proposed method is a self-starting, unconditionally stable and second-order accurate implicit algorithm without overshoot. Particularly, the second-order accuracy in time is achieved in its final form, but it is not required in each sub-step. Its unique algorithmic parameter is analyzed to achieve the unconditional stability and it shares the identical effective stiffness matrix inside three sub-steps to save the computational cost in linear analyses. The same as the Bathe algorithm, the proposed algorithm is always L-stable, meaning that the spurious high-frequency modes can be effectively eliminated. Three numerical examples are simulated to illustrate the superiority of the proposed algorithm over some existing implicit algorithms. The first numerical simulation, solving a linear single-degree-of-freedom system, shows less period elongation errors and the second-order accuracy of the present scheme. The second one, a clamped-free bar excited by the end load, shows the ability of effectively damping out the unexpected high-frequency modes. The last example solves the nonlinear mass-spring system with variable degree-of-freedoms and illustrates that the composite sub-step algorithm can save more computational cost than the traditional implicit algorithm when the integration step size is selected appropriately.     

基于小波变换和改进萤火虫算法优化极限学习机的短期负荷预测

提出了一种基于小波变换和改进萤火虫优化极限学习机的短期负荷预测方法.通过小波分解和重构,对原始负荷序列进行降噪;在模型训练阶段利用改进的萤火虫算法优化极限学习机参数,获得各序列的最优模型;针对各子序列分别预测叠加得到最终预测值.通过在两种时间尺度的数据序列上进行数值计算,与传统的ARMA、BP神经网络、支持向量机及LSSVM等多种经典预测模型相比,模型预测效果更优.     

多层感知器信用评模型及预警研究

本文利用多层感知器 ( MLP)原理建立神经网络信用评价模型 ,用来对我国 2 0 0 0年 1 0 6家上市公司进行信用评级 ,并进一步对我国 2 0 0 1年公布的 1 3家预亏公司进行预警研究 .按照各上市公司的经营状况分为“好”、“差”两类 ,每一类由 5 3家上市公司构成数据样本 .对于每一家上市公司 ,主要考虑其经营状况的四个财务指标 :每股收益 ,每股净资产 ,净资产收益率和每股现金流量 .仿真结果表明 ,本文所建立的神经网络信用评价模型有很高的分类准确率 ,达到 98.1 1 % .又由于该信用评价模型有很强的适应能力 ,故可以进一步用来对企业的财务危机进行预警研究 .预警实证分析表明 ,该信用评价模型对我国 2 0 0 1年公布的 1 3家预亏公司进行预警分析 ,预警准确率达到 1 0 0 % .此外 ,文中还给出 MLP网络模型的学习算法和步骤