International cooperation in assuring continuity of environmental satellite data

In the 25 years since the launch of the first weather satellite, meteorological observations from space have become an essential part of weather forecasting and global environmental monitoring. Beginning in the 1970s, constrained national budgets and the need for a coordinated approach to global satellite observing have caused satellite operators to pursue international cooperation to assure the continuity and compatibility of these systems. This article reviews current bilateral and multilateral cooperation and technical coordination in environmental satellite activities. It also explores the potential for alternative institutional arrangements for maintaining the continuity of environmental satellite data in the decades to come.     

A first step towards proton flux forecasting

We present a preliminary version of a potential tool for real time proton flux prediction which provides proton flux profiles and cumulative fluence profiles at 0.5 and 2 MeV of solar energetic particle events, from their onset up to the arrival of the interplanetary shock at the spacecraft position (located at 1 or 0.4 AU). Based on the proton transportation model by Lario et al. [Lario, D., Sanahuja, B., Heras, A.M. Energetic particle events: efficiency of interplanetary shocks as 50 keV E < 100 MeV proton accelerators. Astrophys. J. 509, 415–434, 1998] and the magnetohydrodynamic shock propagation model of Wu et al. [Wu, S.T., Dryer, M., Han, S.M. Non-planar MHD model for solar flare-generated disturbances in the Heliospheric equatorial plane. Sol. Phys. 84, 395–418, 1983], we have generated a database containing “synthetic” profiles of the proton fluxes and cumulative fluences of 384 solar energetic particle events. We are currently validating the applicability of this code for space weather forecasting by comparing the resulting “synthetic” flux profiles with those of several real events.     

基于支持向量机回归的电力负荷预测研究

不同于传统的基于经验风险最小化的回归方法,支持向量机回归方法基于结构风险最小化准则.与神经网络相比,该方法在解决学习精度和推广性之间的矛盾方面有明显的优势.本文以城市电力负荷预测为应用背景,对比研究了基于统计学习理论的支持向量机回归方法和神经网络方法.预测结果显示支持向量机可能是一种非常有前景的预测工具,其预测精度明显好于神经网络.     

基于可靠性数据的航材备件需求预测方法

航材备件的需求预测已成为装备综合保障领域研究的难点.主要对各部件寿命分布类型进行归纳与总结,建立多数零件服从寿命分布类型的需求预测模型,并对军机执行任务这一特殊情况建立了战损备件预测模型.以可靠性数据为研究基础,对正态分布、指数分布、威布尔分布等备件进行分类分析,根据实际情况逐步精确备件保障概率的取值.计算结果表明,方法能给出合理的备件保障概率和备件需求预测结果.     

基于反馈补偿K-means的救援物资需求预测

针对低空应急救援过程中救援物资需求的多样性与不确定性，提出基于反馈补偿K-means相似搜索的物资需求预测算法。通过比较非平稳数据处理方法，采用差分自回归平均移动模型（ ARIMA）对历史灾情数据进行平滑处理；将预处理后的灾情数据运用基于反馈补偿的K-means方法进行聚类分析；再对比夹角余弦，杰卡德相似系数以及相关系数这3种方法，搜索出相似度最大历史灾情案例，并线性求解当前低空应急救援所需物资量。实验结果表明，在基于反馈补偿K-means相似搜索的物资需求预测过程中，运用相关系数搜索的误差是最小的，方法不仅提高了大数据处理能力，而且在一定程度上提高了预测精度。     

Variations in statistical parameters of the NmF2 equinoctial asymmetry with latitude and solar activity near noon

Hourly values of the F2-layer peak density, NmF2, measured by 95 ionosondes near noon from 1957 to 2011 at low and middle geomagnetic latitudes of the northern and southern geographic hemispheres are used in a statistical study of the NmF2 equinoctial asymmetry. The ratios, R, of NmF2 measured during 61 days around the March equinox to NmF2 measured during 61 days around the September equinox at the same UT near noon during geomagnetically quiet daytime conditions for approximately the same solar activity conditions over the same ionosonde are analyzed. The conditional probability of the occurrence of R in an interval of R, the most probable value of R, and the mean expected value of R are calculated for the first time for the low, moderate, and high solar activity levels to study variations in these statistical parameters with latitude and solar activity. These statistical parameters are averaged over 5° geomagnetic latitude interval in the northern and southern geographic hemispheres to calculate and to study for the first time trends in latitude and solar activity of these averaged NmF2 equinoctial asymmetry statistical characteristics.     

基于支持向量机的航空发动机性能衰退指标预测

针对航空发动机性能衰退状态的趋势预测问题,基于非线性支持向量机回归建立了发动机性能衰退指标的预测模型,给出了反映其性能衰退的综合指标。利用该模型对发动机的性能衰退指标进行了预测分析,并与神经网络模型预测结果进行了比较。结果表明:支持向量机回归预测模型能更准确地对发动机的未来状况进行预测。     

飞行器太阳能源计算方法CSCD

介绍了空间飞行器典型电源系统组成,通过对影响太阳能源输出的各项参数因素分析,提出了空间飞行器太阳能源计算方法,该方法包括由功率计算所需太阳阵面积的公式和由太阳阵面积计算输出功率的公式,达到了准确计算太阳能源的效果,解决了以往空间飞行器太阳能源估算误差大,在总体回路设计中电源参数难以闭环的问题。     

国有企事业单位科技人才需求量预测

本文采用灰色系统理论，对我国国有企事业单位科技人才需求量进行科学的预测，计算出今后12年内每年国有企事业单位对科技人才的需求量，计算结果经检验，模型为一级精度。这为我国发展知识经济、科技进步及高校建设提供了可信的参考数据和模型。     

时序法预测股市的新途径

文章在传统的经典时序法模型的基础上进行了改进,把时序法与技术分析结合起来,并对时序法得出的趋势进行二次时序法处理.从经典时序法对趋势的判断,转为对趋势的变动进行判断,使之更能适应股票市场的多变性,并能够有效的应用于实践,其具体方法是二次时序法和引入反转因子.