一种改进的合成流量法及其在水文预报中的应用

合成流量法对下游站流量作出预报的关键是确定各上游站流量到达下游站的时间。在实际中常采用平均传播时间,然而不同水情下真实传播时间与平均传播时间存在一定的差距,影响预报精度。对此,提出时间窗口概念,以平均传播时间为中心,向前、后各开一个时间窗口,用上游各站时间窗口内流量的线性组合表示其到达下游站的流量,以提高模型对不同水情的适应能力和预报精度。以三峡水库2009~2015年流量数据率定参数,以2016~2018年流量数据作为预报对象,试验结果表明时间窗口法能够显著改善预报精度。     

A review on the features and technologies for energy efficiency of smart grid

In today's world striving for efficiency in every sector, especially power generation and distribution, smart grids emerge as the solution for efficiently meeting the increasing demand. They adjust themselves to optimally deliver energy at the lowest cost and highest quality possible. The grid successfully makes use of renewable energy resources, electric vehicles, and smart pricing techniques in its attempt to achieve energy efficiency. It also promotes a greener environment by striving to reduce greenhouse gas emissions. Information communication technology (ICT) helps the grid in collecting consumption data from the consumers and in sharing tariff information. ICT also helps to gather information about the status of the grid with regard to aspects like power quality, faults etc. The purpose of this paper is to review recent literature with a view to comprehensively present the technologies employed in the smart grid for achieving energy efficiency and the challenges involved therein.     

基于数据驱动理念的电力日负荷曲线预测方法研究

针对短期日负荷预测的精度问题,本文提出一种基于数据驱动理念的电力负荷预测方法。在建立预测模型前对所给数据采取一定的预处理:首先提取所收集的海量数据的负荷特征,对负荷特征进行分析,然后进行负荷数据与影响负荷值的因素之间的相关性分析,以此确定对负荷影响较密切的因素,随后建立分类器得到各主要影响因素与各负荷类别之间的关系为后续预测模型奠定基础。对预处理后得到的不同类型的负荷数据采用最小二乘支持向量机方法建立不同的负荷预测模型。以南方某发达城市2008年的负荷数据作为算例验证数据,将本文所提负荷预测方法所得结果与未经数据预处理的负荷预测方法所得结果进行比较,结果表明本文提出的方法得到的预测结果精度较传统方法提高约6%。     

鲸鱼优化支持向量机的短期风电功率预测

为提高风电预测的精度,提出一种鲸鱼优化支持向量机SVM(support vector machine)的组合预测模型。该模型针对风电序列的非平稳波动特性,首先应用集合经验模态分解技术EEMD(ensemble empirical mode de?composition)将原始风电序列分解为一系列不同特征尺度的子序列;并引入鲸鱼优化算法WOA(whales optimiza?tion algorithm)解决SVM中学习参数选择难的问题,进而对各子序列建立WOA_SVM预测模型;最后,叠加各子序列的预测值以得到最终预测值。仿真表明,所提EEMD_WOA_SVM模型具有较高的风电预测精度,显著优于其他基本模型。     

基于SARIMA-GRNN-SVM的短期商业电力负荷组合预测方法

针对短期商业电力负荷预测准确性与周期难以满足现有电力现货市场的问题,提出了一种基于SARIMAGRNN-SVM(seasonal autoregressive integrated moving average-generalized regression neural network-support vector machine)的商业电力负荷组合预测模型。首先,对商业电力负荷变化的周期规律与随机因素的复杂影响进行了分析;然后,结合以上分析,选用SARIMA和GRNN为单一预测模型对商业电力负荷进行预测,并利用SVM进行组合,实现日前商业电力负荷预测;最后,通过某商业综合体的电力负荷数据进行验证。所提组合预测模型较单一预测模型拥有更优的预测精度与鲁棒性,可以为短期商业电力负荷预测提供借鉴。     

A new interval diagnosis method: Application to the spacecraft rendezvous phase of the Mars sample return mission

This paper presents a new optimized interval principal component analysis applied to detect and isolate actuators faults of an autonomous spacecraft involved in the rendezvous phase of the Mars sample return mission. Based on the exploitation of various arithmetic and interval analysis properties, the new interval model is built by solving the interval eigenpairs problem via a resolution of a parametric linear programming problem. The detection and isolation phases are performed by extending the classic methods to interval-valued data. The proposed method is applied to detect and isolate actuators faults that can occur on the spacecraft's thrusters. Based on data provided by a “high fidelity” industrial simulator developed by Thales Alenia Space, the obtained results proved the effectiveness of the proposed interval fault diagnosis method on detecting and isolating thrusters' faults.     

Estimation of a preference map of new consumers for spatial load forecasting simulation methods using a spatial analysis of points

The paper presents a spatial analysis of points especially suited to estimate a preference map for new consumers, which is then used as an analytical tool in spatial electric load forecasting. This approach is an exploratory spatial data analysis used to discover useful point patterns in the spatial location of distribution transformers to calculate a preference value for each area, rating it with respect to a hypothetical load change that may occur. We consider the locations of distribution transformers occupied land. Random points are generated in the study area where the new loads are expected; these points are referred to as unoccupied land. The method uses a generalized additive model (GAM) to estimate the probability of unoccupied land becoming occupied land. We test the approach with data from a real distribution system in a mid-size city in Brazil; the result is a preference map that shows the areas where new consumers are most likely to be allocated. The main advantage of this method is the ability work with a small-scale resolution, which enables the use of a resolution suitable for spatial load forecasting method chosen. We test the calculated probabilities in a spatial load forecasting simulation, yielding results with lower spatial error when compared with the heuristic technique.     

Seasonal hydrological and nutrient loading forecasts for watersheds over the Southeastern United States

We show useful seasonal deterministic and probabilistic prediction skill of streamflow and nutrient loading over watersheds in the Southeastern United States (SEUS) for the winter and spring seasons. The study accounts for forecast uncertainties stemming from the meteorological forcing and hydrological model uncertainty. Multi-model estimation from three hydrological models, each forced with an ensemble of forcing derived by matching observed analogues of forecasted quartile rainfall anomalies from a seasonal climate forecast is used. The attained useful hydrological prediction skill is despite the climate model overestimating rainfall by over 23% over these SEUS watersheds in December–May period. The prediction skill in the month of April and May is deteriorated as compared to the period from December–March (zero lead forecast). A nutrient streamflow rating curve is developed using a log linear tool for this purpose. The skill in the prediction of seasonal nutrient loading is identical to the skill of seasonal streamflow forecast.     

An optimization decision support approach for risk analysis of carbon emission trading in electric power systems

Concerns over dramatic increasing electricity demand, exacerbating power shortage and changing climatic condition are emerging associated with municipal electric power systems (EPS). In this study, a risk-explicit mixed-integer full-infinite programming (RMFP) approach is developed for planning carbon emission trading (CET) in EPS. RMFP-CET has advantages in risk reflection and policy analysis, particularly when the input parameters are provided as crisp and functional intervals as well as probabilistic distributions. The developed method is applied to a real case study of CET planning of EPS in Beijing. Various electricity policies are incorporated within the modeling formulation for enhancing the RMFP-CET's capability. The results indicate that reasonable solutions have been generated, which are useful for making decisions of electricity production and supply as well as gaining insight into the tradeoffs among electricity supply risk, system cost, and CO₂ mitigation strategy.     

An inventory model with truncated exponential replenishment intervals and special sale offer

In this study, we develop an inventory model with stochastic replenishment intervals and special sale offer from a supplier. The replenishment interval is assumed to obey a truncated exponential distribution and shortage is partially backordered. Our goal in this research is to maximize the total profit of cost savings due to special sale offer from supplier. A closed-form solution of the model and its convexity condition is developed. A numerical example with real world data is provided to illustrate the theory.