Generalized weighted functional proportional mean combining forecasting model and its method of parameter estimation

A new kind of combining forecasting model based on the generalized weighted functional proportional mean is proposed and the parameter estimation method of its weighting coefficients by means of the algorithm of quadratic programming is given. This model has extensive representation. It is a new kind of aggregative method of group forecasting. By taking the suitable combining form of the forecasting models and seeking the optimal parameter, the optimal combining form can be obtained and the forecasting accuracy can be improved. The effectiveness of this model is demonstrated by an example.

Average Power Function of Noise and Its Applications in Seasonal Time Series Modeling and Forecasting

This paper presents a new method of detecting multi-periodicities in a seasonal time series. Conventional methods such as the average power spectrum or the autocorrelation function plot have been used in detecting multiple periodicities. However, there are numerous cases where those methods either fail, or lead to incorrectly detected periods. This, in turn in applications, produces improper models and results in larger forecasting errors. There is a strong need for a new approach to detecting multi-periodicities. This paper tends to fill this gap by proposing a new method which relies on a mathematical instrument, called the Average Power Function of Noise (APFN) of a time series. APFN has a prominent property that it has a strict local minimum at each period of the time series. This characteristic helps one in detecting periods in time series. Unlike the power spectrum method where it is assumed that the time series is composed of sinusoidal functions of different frequencies, in APFN it is assumed that the time series is periodic, the unique and a much weaker assumption. Therefore, this new instrument is expected to be more powerful in multi-periodicity detection than both the autocorrelation function plot and the average power spectrum. Properties of APFN and applications of the new method in periodicity detection and in forecasting are presented.

Function S-rough sets and two law forecast

Function S-rough sets has the properties of dynamics, heredity, and memory. Function S-rough sets is penetrated and crossed with the issue of economic law forecast, then a new forecast model based on function S-rough sets namely the two law forecast model is proposed, which includes upper law forecast model and lower law forecast model; and its＇ implement algorithm is given. Finally, the validity of the model is demonstrated by the forecast for region economic development of Hainan Province.

Sensitivity Analysis of Radial Basis Function Networks for River Stage Forecasting

Sensitivity analysis of neural networks to input variation is an important research area as it goes some way to addressing the criticisms of their black-box behaviour. Such analysis of RBFNs for hydrological modelling has previously been limited to exploring perturbations to both inputs and connecting weights. In this paper, the backward chaining rule that has been used for sensitivity analysis of MLPs, is applied to RBFNs and it is shown how such analysis can provide insight into physical relationships. A trigonometric example is first presented to show the effectiveness and accuracy of this approach for first order derivatives alongside a comparison of the results with an equivalent MLP. The paper presents a real-world application in the modelling of river stage shows the importance of such approaches helping to justify and select such models.

TRIZ TECHNOLOGY FORECASTING AS QFD INPUT WITHIN THE NPD ACTIVITIES

As a result of the fierceness of business competition, companies, to remaincompetitive, have to charm their customers by anticipating their needs and being able to rapidlydevelop exciting new products for them. To overcome this challenge, technology forecasting isconsidered as a powerful tool in today's business environment, while there are as many successstories as there are failures, a good application of this method will give a good result. Amethodology of integration of patterns or lines of technology evolution in TRIZ parlance ispresented, which is also known as TRIZ technology forecasting, as input to the QFD process to designa new product. For this purpose, TRIZ technology forecasting, one of the TRIZ major tools, isdiscussed and some benefits compared to the traditional forecasting techniques are highlighted. Thena methodology to integrate TRIZ technology forecasting and QFD process is highlighted.

Application test of matter element analysis in earthquake forecast

Calculation by means of the previous indices of the seismic activity can have the matter element analysis possess the forecast function. Readjusting repeatedly the grade limit value of every index can maximize the historical fitting ratio of the calculated and actual grade of the annual maximum magnitude, whose result is relatively ideal.

Modeling Breast Cancer Incidence Rates: A Comparison between the Components of Functional Time Series (FTS) Model Applied on Karachi (Pakistan) and US Data

Several studies showed that the breast cancer incidence rates are higher in high-income (developed) countries, due to the link of breast cancer with several risk factors and the presence of systematic screening policies. Some of the authors suggest that lower breast cancer incidence rates in low-income (developing) countries probably reflect international variation in hormonal factors and accessibility to early detection facilities. Recent studies showed that the breast cancer increased rapidly among women in Pakistan (a developing country) and it became the first malignancy among females of Pakistan. Although, the incidence rates may contain important evidence for understanding and control of the disease;however in Pakistan, the breast cancer incidence data have never been available in the last five decades since independence;rather, only hospital-based data are available. In this study, we intend to apply Functional Time Series (FTS) models to the breast cancer incidence rates of United State (developed country), and to see the difference between various components (age and time) of Functional Time Series (FTS) models applied independently on the breast cancer incidence rates of Karachi (Pakistan) and US. Past studies have already suggested that the incidence of US breast cancer cases was expected to increase in the coming decades. A progressive increase in the number of new cases is already predetermined by the high birth rate that occurred during the middle part of the century, and it will lead to nearly a doubling in the number of cases in about 4 decades. We also obtain 15 years predictions of breast cancer incidence rates in United States and compare them with the forecasts of incidence curves for Karachi. Development of methods for cancer incidence trend forecasting can provide a sound and accurate foundation for planning a comprehensive national strategy for optimal partitioning of research resources between the need for development of new treatments and the need for new research directed toward primary preventive measures.

Inventory Management and Demand Forecasting Improvement of a Forecasting Model Based on Artificial Neural Networks

Forecasting is predicting or estimating a future event or trend.Supply chains have been constantly growing in most countries ever since the industrial revolution of the 18th century.As the competitiveness between supply chains intensifies day by day,companies are shifting their focus to predictive analytics techniques to minimize costs and boost productivity and profits.Excessive inventory(overstock)and stock outs are very significant issues for suppliers.Excessive inventory levels can lead to loss of revenue because the company's capital is tied up in excess inventory.Excess inventory can also lead to increased storage,insurance costs and labor as well as lower and degraded quality based on the nature of the product.Shortages or out of stock can lead to lost sales and a decline in customer contentment and loyalty to the store.If clients are unable to find the right products on the shelves,they may switch to another vendor or purchase alternative items.Demand forecasting is valuable for planning,scheduling and improving the coordination of all supply chain activities.This paper discusses the use of neural networks for seasonal time series forecasting.Our objective is to evaluate the contribution of the correct choice of the transfer function by proposing a new form of the transfer function to improve the quality of the forecast.

变形监测灰色预测模型对比及替代方法研究

将变形监测灰色预测模型分为传统GM(1,1)模型及其改进模型、非齐次灰色模型、GM(1,1)幂模型及其改进模型3种类型,以Origin拟合函数Exp2PModl、Exponential和SRichards2作为3类灰色预测模型的替代方法,基于理论研究和实例验证对比分析3类灰色预测模型及其替代方法。结果表明,3类灰色预测模型在拟合函数、有无极限值、适合等时距或非等时距建模和适用范围等方面存在显著差异,需要根据变形监测数据特征选择合适的灰色预测模型类别;与3类灰色预测模型相比,Origin拟合函数在参数求解和建模数据要求上更具优势,而且可以得到相当甚至更高的拟合或预测精度,除需要编程实现的特殊优化目标外,完全可以代替灰色预测模型用于变形监测。

计及预测误差时变相关特性的新型电力系统爬坡容量需求分析方法

为提高新型电力系统爬坡容量需求分析的准确性,提出了计及净负荷预测误差时变相关特性的爬坡容量需求分析方法。首先明确新型电力系统所需要的爬坡容量构成,分析爬坡需求与净负荷预测误差之间的关系。然后,建立了净负荷预测误差与净负荷预测值的动态Copula函数,通过演进方程构建Copula函数参数的时序关联关系,应用条件概率理论建立由于净负荷预测误差而产生的爬坡需求概率分布模型,以置信区间对爬坡需求概率分布结果进行离散化表征,从而得到相应的爬坡容量需求。最后,基于我国西部某电网的实际运行数据,以是否考虑预测误差时变相关特性构建了3种不同的测算模型,算例计算结果验证了所提方法的有效性和正确性。

基于改进扩散模型的温度预报

针对传统数值预报模式计算时间长和计算资源消耗大的问题,以及现有深度学习预报方法在温度预报结果上不精确,且预测结果模糊的问题,提出了一个新的温度预报模型。首先,设计了一个时空信息捕捉模块,将该模块捕获的长期依赖信息,作为扩散模型的生成条件,赋予扩散模型预报的能力;其次,设计了一个新的平衡损失函数,同时保护了扩散模型的生成能力和时空信息捕捉模块对时空信息的捕捉能力;最后,基于美国国家环境预报中心的再分析数据进行预报,与现有的深度学习方法相比,所提模型预报结果的质量在均方误差(mean square error,MSE)上降低了17.3%,在均方根误差(root mean square error,RMSE)上降低了9.14%,在峰值信噪比(peak signal to noise ratio,PSNR)上提升了5.1%。改进的扩散模型能有效地捕捉时空依赖的关系,有效地进行时空序列预测,效果优于其他对比方法。

基于改进Transformer时序算法的区域经济预测

现有的区域经济预测模型指标变量存在冗余,且忽略了行业、区域等静态变量对预测结果的影响,导致预测效率不高.针对上述问题,提出了基于改进Transformer时序算法的区域经济预测模型.首先利用Copula函数对传统Transformer模型进行优化(Coformer);其次选取区域经济的影响指标,对其进行主成分分析,去除冗余信息;然后将降维后的指标变量和静态变量作为Coformer的输入,对变量进行编码,并通过多头注意力机制增强重要信息,最后用解码器对编码的变量解码,利用Softmax输出区域历年生产总值序列的预测结果.实验结果表明,所提模型的预测准确率为0.908,比另外三种模型分别提高了15.9%、12.3%和6.7%,表现出了优异的预测性能.

基于区域空间经济整体规划与综合运输超级网络的铁路货运量预测建模方法研究

研究以PECAS整体规划建模框架为基础,提出基于区域空间经济整体规划与综合运输超级网络的铁路货运量预测建模方法。首先,运用PECAS模型实现社会经济活动时空变迁及其影响下的大区域内货运生成与分布预测。其次,以综合运输超级网络为基础,通过构建包括公、铁、水多方式的货运广义阻抗函数来模拟货物运输方式之间的竞合关系及承运人的运输路径选择行为。最后,基于随机用户均衡实现综合运输超级网络的一体化网络货流分配。研究以长江经济带为例,对区域铁路货运需求进行预测,将网络货流分配后主要断面流量的预测值与观测值进行对比,以检验模型的准确性。研究结果表明,所提出的方法在大区域范围内可以较好地预测铁路货运流量的分配情况,为区域内铁路货运需求预测与铁路基础设施规划提供决策支持。

基于LSTM神经网络的基坑工程智能预警系统研发与应用

基坑开挖过程中伴随有支护结构及周围岩土体的受力和形变状态的改变,因此在工程建设中对基坑进行监测十分必要。为解决基坑监测智能化程度低、可视化程度低、预警精确度较低导致频繁报警、监测数据更新共享速度慢等问题,采用B/S模式、Vue前端、C#语言后端、SQLServer2012数据库等并嵌入python语言编写长短期记忆(Long Short-Term Memory,LSTM)神经网络算法模型开发一套基坑智能预测预警系统。该系统实现了信息集中管理、数据存储与查看、数据算法自动计算、自动绘制图表、自动报警预警、快速生成报警报告等功能。通过在苏州某地铁基坑开挖过程的应用,证明了本系统能够综合利用监测预警与基于LSTM神经网络模型的超前预测预警两种预警模式为施工人员准确掌握基坑开挖过程中支护结构及周围土体变形情况提供技术支持与保障,具有很强的现实使用意义。

基于水电储能调节的风光水发电联合优化调度策略

为缓解新能源装机容量扩大引起的弃风弃光现象,在已有梯级水电上下电站之间加入储能泵站,提出风光水储短期优化调度策略。构建以风光水储系统负荷跟踪误差最小、梯级水电站发电量最大和梯级水电站发电耗水量最小的多目标优化调度模型;提出基于季节性自回归移动平均(seasonal auto-regressive lntegrated moving average, SARIMA)模型和Copula函数的风光出力预测模型作为优化调度模型的边界条件,通过SARIMA预测模型将风光出力历史数据分解为季节性分量、趋势分量以及随机噪声余项进行全天96个调度时段风光出力预测,并叠加上基于Copula函数生成风光出力预测误差,然后通过拉丁超立方采样以及K-means聚类进行场景生成和缩减得到5个风光出力场景。选取风光典型日出力数据为例进行算例分析,算例结果表明:所提预测模型较SARIMA模型可以显著提高预测准确度,模型预测风光出力均方根误差从33.34、229.49 MW分别下降至0.697、9.534 MW;所提优化调度策略可以在全年丰、平、枯水期有效减少弃风弃光现象,并可将过剩新能源中的50%转化为上级水库储存水能。

基于GA-BP神经网络模型城市河道水位预报研究

城市内河水位预报对城市内涝风险管理具有重要意义。传统数值模拟模型计算效率较低,且无法实时计算。针对以上问题,提出一种基于Gaussian函数改进BP神经网络的河道水位预报模型,解决了BP神经网络模型预报精度低、在误差平坦区收敛速度慢的问题。该方法利用Gaussian函数改进BP神经网络梯度下降算法,针对模型不同权重与阈值设定不同学习率,对各参数进行针对性优化,能够有效加速BP神经网络模型训练效率;针对模型在误差平坦区收敛速度慢的问题,通过Gaussian函数增大梯度下降算法在误差平坦区的学习率,控制梯度下降算法在误差较大时的学习率,能够有效加速BP神经网络模型在误差平坦区的收敛速度。以福州市晋安区6个河道水位测站为研究对象,构建GABP神经网络河道水位预报模型进行城市内河水位预报,并探讨不同降雨输入形式对河道水位预报精度的影响。结果表明:GA-BP神经网络能够有效提升BP神经网络在误差平坦区的收敛速度与模型预报精度,试验集预报纳什效率系数(NSE)均在0.8以上,能够将预报峰值水位相对误差控制在5%以内,其中降雨以小时降雨量形式输入能够将预报NSE提升至0.9以上。研究表明采用Gaussian函数改进BP神经网络模型能够有效提升模型预报精度,对提升城市河道水位预报具有重要意义。

Copula分位数回归方法在风电超短期出力预测上的应用

风电出力具有较强的随机性和波动性,相比于传统预测,分位数预测方法能够提供全面的风电功率概率分布信息,可实现更可靠的风电出力预报,对电网系统的安全和稳定运行具有重要意义.以甘肃某风电站为案例,将数据按6∶2∶2划分为训练集、验证集和测试集,采用基于Copula的分位数回归方法(QCopula)进行功率区间预测,并与三个传统的分位数回归方法进行比较.结果显示,在不同置信区间下QCopula的修正预测区间精度范围在0.701~0.773之间,预测精度平均值比传统分位数回归(QR)、随机森林分位数回归(QRF)和长短期记忆神经网络分位数回归(QLSTM)分别高出15%、9%和13%,优于其他三种分位数预测方法.分位数交叉验证中,QCopula未出现分位数交叉,每个样本点的功率预测值均随概率值单调递增,而QR、QRF、QLSTM均出现不同程度的分位数交叉现象.综上所述,QCopula可以表征更小的区间宽度和更高的区间覆盖率,且分位数曲线不存在交叉,可信度较高.

售电侧开放市场环境下基于多分位鲁棒极限学习机的短期负荷预测技术

该文基于极限学习机算法设计一种用于短期负荷预测的多分位鲁棒极限学习机模型,该模型能解决传统预测模型抗干扰能力差的缺陷,可以在面临不确定性因素干扰的情况下准确预测负荷。对传统模型和多分位鲁棒极限学习机模型的鲁棒性和多分位性进行验证,对比结果表明,多分位鲁棒极限学习机模型的鲁棒性更好,在不同分位下的预测精度更高。

基于极大熵法的三峡水库洪水预报误差分布规律

深入研究洪水预报误差的分布规律具有很大的实用价值,可为水库调度决策的拟定提供辅助参考。通常的研究方法为正态分布法与对数正态分布法,但这类研究方法受人为主观性的影响较大。使用三峡水库2020-2022年入库洪水数据将洪水分为4个量级,建立三峡水库洪水洪量预报误差分布的极大熵模型,得出不同预见期下4个量级的洪水洪量预报相对误差的分布规律,结果表明:采用极大熵模型可以定量地获得三峡水库的洪水预报误差分布规律,具有较强的适用性且可进一步推广使用。

“双碳”目标下中国能源需求未来走势预测分析

“双碳”目标下准确掌握能源需求对社会经济发展至关重要。文章采用灰色相对关联度对各变量间的相关性进行定量刻画,并诊断自变量是否存在明显多重共线性;运用逐步回归分析法精细化分析能源需求影响因素,挖掘能源需求有效驱动因素,构建“最佳”能源需求函数模型。结果表明:(1)能源价格、消费水平和经济发展水平是中国能源需求的有效驱动因素。(2)基于有效影响因素建立的“最佳”逐步回归双对数能源需求函数模型具有良好的预测性能,可用于中国能源中长期需求预测,预测结果可作为科学制定能源政策的重要参考依据。(3)未来20年中国能源需求增长量主要来源于清洁能源需求量的增长,到2040年中国能源需求量约62.6亿吨标准煤,年均增长率约0.8%。