改进的灰色模型在港口吞吐量预测中的应用

用提高原始数据列光滑度的方法来提高灰色预测模型(GM模型)的精度。利用函数a-xm(a>1,m>1)对原始离散数据列｛x(0)(k)｝进行变换,原始数据列变成a-[x(0)(k)]m ,很大程度上提高了预测精度。将这种方法应用于港口吞吐量的预测,结果比传统灰色预测方法精度高很多,对正确做出港口发展战略、统筹安排、正确决策以及减少损失有重大意义。     

基于聚类分析的港口类型化研究

根据各港口集装箱运量增长的特点,通过软件SPSS,应用基于聚类分析的类型化方法将我国沿海主要集装箱港口分成三大类型,即普通增长型、加速增长型和波动增长型,并对各类型港口增长特点进行了分析.     

利用回归预测技术进行港口吞吐量预测的方法研究

以某港口近15年的货物吞吐量作为原始数据,采用回归分析技术,建立回归预测模型,得到预测结果,为港口的发展战略研究提供一定的技术支持。     

Telecommuting and travel: state of the practice,state of the art

This paper provides an overview of the status of telecommuting in the United States, especially as it relates to changes in travel behavior. Regarding the state of the practice, the paper discusses some refinements to the definition of telecommuting that have developed through increased operational experience. It reports several policy statements involving telecommuting, and explores the appeal of telecommuting as a public policy instrument. It highlights some trends in the implementation of home-based and work center-based telecommuting, and suggests that visible public-sector involvement has been crucial to the increased activity in this area.In sketching the state of the art, the paper outlines some frequently-stated hypotheses on telecommuting and travel behavior, and summarizes current empirical findings relating to those hypotheses. Finally, it suggests a variety of topics suitable for further research. These include studying factors influencing the ultimate adoption levels of telecommuting; impacts on energy/air quality, mode choice, and location/urban form; interactions with other transportation demand management strategies; relationships to the traditional urban travel demand forecasting process; cost/benefit tradeoffs; and telecommuting centers.     

齐岳山隧道F12断层注浆加固技术

注浆技术是通过隧道不良地质的常用方法之一。齐岳山隧道穿越多个断层地段,通过长、中、短超前地质预测预报技术,准确判定隧道围岩地质,施工中采用注浆回固技术通过断层。     

基于多次累加和函数拟合的 结构初期沉降预测

传统沉降预测方法对于量值小、相对波动大 的短期沉降不适用,为此提出适用于结构沉降反复剧 烈波动的自修正短期动态预测方法。该方法首先对已 获取的第 1 ～ n 时刻的结构沉降数据进行多次累加,得 到一条单调光滑数据累加曲线,然后对该曲线进行最 高( n － 1) 次多项式拟合,外延得到第( n + 1) 时刻的沉 降预测累加值,再通过回归递推,即得到第( n + 1) 时刻 的沉降预测值。同时,对该方法引入“新陈代谢”思想, 以更好地反映沉降变化趋势的目前特征。最后,将某 盾构施工现场结构沉降预测值与实测值进行比较。结 果表明,该方法能取得较好的预测结果。     

高速公路交通流的分形维数与相空间重构预测

对成渝高速公路短时交通流通过计算不同时间尺度下Hurst指数而等到其相应的分形维数,结果表明,时间间隔越短的交通流,其分形维数越大,结构越复杂.由于时间间隔越短的交通流随机性大和复杂的结构,所以预测也就越困难.提出了一种新的基于相空间重构和移动平均相结合的预测方法——移动平均最近邻域法,从理论与实际数据两方面分析和验证了该方法对短时交通流预测的有效性.     

城市轨道交通供电系统负载特征分析

通过对上海轨道交通1号线供电系统主变电站、牵引变电站和降压变电站的负载特征及典型变电站的负载随季节性变化的特征进行分析,验证了典型变电站的日负载曲线的季节变化特征与该线路供电系统的月度用电量数据的关联性。分析结果可为城市轨道交通线路的负载预测及新线设计过程中的负载模型的仿真提供数据参考和分析方法。     

基于RBF神经网络的铁路沿线短时风速预测方法

对实测风速数据进行Kalman滤波,去除实测风速数据的偏差;通过归一化处理,消除数据中的冗余成分;针对RBF神经网络的预测误差会随着时间的推移而增大的问题,采用滚动式训练方法在线训练RBF神经网络;用训练好的RBF神经网络进行风速预测,再对预测结果进行反归一化处理,得到最终的预测风速.仿真结果表明,运用基于RBF神经网络的铁路短时风速预测方法对短时风速进行预测,最大相对误差仅为5.92％,可满足铁路防灾安全监控系统中风速预测子系统的要求.     

Impact of Sophisticated Stationary Forecast Techniques on the Bullwhip Effect in a Supply Chain

The bullwhip effect in a multistage supply chain was analyzed using sophisticated stationary forecasts (third order moving average and third order exponential smoothing forecasts). The third order exponential smoothing and third order moving average forecasts sometimes have a variance reducing effect on the supply chain.In a supply chain with positively correlated or independent and identically distributed (i.i.d) demands, the order variance based on simple moving average forecast (or simple exponential smoothing forecast) is larger than the order variance based on second order moving average forecast (or second order exponential smoothing forecast),and the order variance based on second order moving average forecast( or second order exponential smoothing forecast) is larger than the order variance based on third order moving average forecast( or third order exponential smoothing forecast). In addition, for i.i.d demands, third order exponential smoothing forecast leads to a larger variation than third order moving average forecast.