基于广义回归神经网络的瓦斯涌出量预测

简要介绍了瓦斯涌出量预测问题和广义回归神经网络（GRNN）的特点,指出与常用的BP神经网络相比,使用广义回归神经网络（GRNN）具有收敛迅速、人为干扰小等优点,适宜用于瓦斯涌出量的预测。并对一个案例进行预测,证明了广义回归神经网络（GRNN）可以满足实际生产的精度要求,较好解决瓦斯涌出量预测的问题。     

长江经济带沿江铁路通道规划建设必要性分析

长江经济带发展正式上升为国家战略。沿江铁路通道横跨我国长江经济带,是我国长江经济带发展战略的支撑性通道。以沿江通道区位特征、产业布局、经济特征为依据,通过分析沿江铁路通道的发展现状、存在问题,通道内不同时期铁路运输需求,以及对拟建项目的功能定位进行分析,从而提出沿江铁路通道规划建设的必要性。     

城市轨道交通客流预测宏观指标统计分析这篇中文摘要写得很空泛，不太容易看出论文要点

城市轨道交通客流预测作为需求分析的有效技术手段,其预测结果的可信度和有效性将直接影响决策的精准度,重要性不言而喻。通过对北京、上海、广州、深圳、成都、南京等20余座城市的轨道交通现状运营数据进行全面整理与归纳,系统阐述网络客流、线路客流、车站客流的诸多特征,从负荷强度、网络平均乘距、线路平均运距、换乘系数、断面高峰小时系数、断面不均衡性、换乘客流量级分布、车站超高峰系数等客流预测关键技术指标进行特征探讨与规律总结,以期协助模型工作者更好地把握预测结果的合理性。     

包海通道西安至安康高速铁路运输需求分析

研究国家高速铁路网包海通道西安至安康高速铁路的运输需求,可为本线主要技术标准、运输组织和建设方案的确定提供决策依据。首先通过旅客运输径路比较分析,确定本线主要承担我国蒙西、陕北、山西、北京、东北地区、关中地区与陕南、西南、广西、海南等地的旅客交流,兼顾陕南重镇安康等与关中城市群、陕北地区的城际客流,其次基于"四阶段法"基本原理,对区域铁路客运量进行分配预测,并按照趋势运量、转移运量、诱增运量进一步分析本线客流密度,最后研究本线客流构成、所在通道的综合交通方式结构,并综合确定本线客车开行方案。     

基于周期时变特点的城市轨道交通短期客流预测研究

分析了城市轨道交通客流的周期时变性特征,并根据该特征在GM(1,1)灰色预测模型的基础上改进了马尔科夫算法,以适用于城市轨道交通短期客流预测。用无偏GM(1,1)模型拟合系统的发展变化趋势,再以此为基础进行了马尔科夫链预测,并采用多转移矩阵排除客流数据中噪声数据的扰动。试验结果表明,改进后的模型在城市轨道交通客流短期预测中具有良好的精确性。     

浅谈TRT6000在黄土隧道中应用技术

结合新建大西铁路乔家山隧道黄土围岩超前预报实例,详细阐述TRT6000超前地质预报系统适用性、原理及操作要点。根据现场检测实例,验证其在黄土隧道中的实用性。     

Total cost analysis: An alternative to benefit-cost analysis in evaluating transportation alternatives

This paper explains the theory in support of total cost analysis (TCA) to compare transportation system alternatives. The full costs of each alternative are first aggregated, including travel time costs and monetizable environmental and social costs. Many costs which are considered on the benefits side of the equation in benefit-cost analysis (BCA) as "cost savings" are brought over to the costs side. Total cost differences among alternatives are then traded off against their estimated non-monetized benefits or impacts, just as a consumer trades off product quality against cost before deciding which product he or she will buy. One advantage of TCA over traditional BCA is that the concept of "total cost" is more easily understood by the public and by political decision-makers than BCA concepts such as "net present worth", "benefit-cost ratio" and "internal rate of return". A second advantage is that there is no suggestion that all "benefits" have been considered; decision-makers are free to use their own value judgements to trade off total cost against non-monetizable social, environmental and economic impacts, just as they trade off quality and convenience against cost when purchasing goods and services in their roles as consumers. The TCA approach is demonstrated in this paper through a case study of two systemwide alternatives for the Baltimore, MD urban area.     

Uncertain demand,modal competition and optimal price-capacity adjustments in air transportation

A new microeconomic model for the operation of an airline facing modal competition with uncertain total demand is developed to analyze optimal price capacity combinations. The novelty is the treatment of the capacity restriction, which is not viewed as affecting negatively individual preferences (e.g. probability of a full flight), but does influence aggregate utility. A mode choice model is used to represent unrestricted individual preferences assuming full availability (phone call demand); air capacity is treated as a variable that acts on the actual choice set. Restricted choices and total demand stochasticity are integrated in welfare calculations (users' benefits and profits). Numerical examples are given and results are analyzed in terms of load factors fare levels, and sensitivity to the stochasticity of requests.This research was partially funded by FONDECYT, Chile, Direction Génerale de l'Aviation Civile, France, the Andes Foundation and the Fulbright Commission.     

高速公路可变收费标准模型研究

在介绍可变收费的理论基础上,分析影响高速公路可变收费标准制定的主要因素,并讨论不同时段收费费率同交通量的关系。在设定的可变收费方案下,建立了出行者选择的二元logit模型,根据调研样本,采用最小二乘法对模型参数标定并对设定的可变收费方案进行评价,结果证明可变收费能使交通流在各时段合理分布,从而提高了高速公路的服务水平。     

Spatiotemporal influence of land use and household properties on automobile travel demand

Understanding the patterns of automobile travel demand can help formulate policies to alleviate congestion and pollution. This study focuses on the influence of land use and household properties on automobile travel demand. Car license plate recognition (CLPR) data, point-of-interest (POI) data, and housing information data were utilized to obtain automobile travel demand along with the land use and household properties. A geographically and temporally weighted regression (GTWR) model was adopted to deal with both the spatial and temporal heterogeneity of travel demand. The spatial-temporal patterns of GTWR coefficients were analyzed. Also, comparative analyses were carried out between automobile and total person travel demand, and among travel demand of taxis, heavily-used private cars, and total automobiles. The results show that: (I) The GTWR model has significantly higher accuracy compared with the Ordinary Least Square (OLS) model and the Geographically Weighted Regression (GWR) model, which means the GTWR model can measure both the spatial and temporal heterogeneity with high precision; (II) The influence of built environment and household properties on automobile travel demand varies with space and time. In particular, the temporal distribution of regression coefficients shows significant peak phenomenon; and (III) Comparative analyses indicate that residents’ preference for automobiles over other travel modes varies with their travel purpose and destination. The above findings indicate that the proposed method can not only model spatial-temporal heterogeneous travel demand, but also provide a way to analyze the patterns of automobile travel demand.