����֧���������ع�ĺ�������Ԥ���о�

针对航班延误难以预测的问题,采用支持向量机回归方法建立航班到港延误预测模型。首先,采用相空间重构理论计算到港延误的延迟时间、嵌入维数和最大Lyapunov指数,发现到港延误时间序列存在混沌特性;将航班到港延误时间序列进行相空间重构,并结合执飞该航班的航空器在上游机场的离港延误构建模型的输入向量;其次,将粒子群算法、差分进化算法和遗传算法进行比较,用于选择最优的模型参数,实验表明,差分进化算法能够以较高概率获得最优的预测模型;最后,比较该模型、单一因素预测模型和相关向量机预测模型的航班延误预测性能。结果表明,该模型的预测性能明显优于另外两种模型,能够有效预测航班延误。     

数学回归法在大孔隙沥青混合料级配设计中的应用

透水型沥青路面(OGFC)是一种具有连通空隙的开级配沥青路面材料,具有排水、降噪等良好的功能性.大孔隙沥青混合料的空隙率直接影响到沥青混合料的路用性能和功能性能,而控制OGFC空隙率主要靠确定合理的关键筛孔通过率,但是对于二点通过率与空隙率的具体关系尚无深入研究.文中采用二元线性回归找出OGFC-13的2个关键筛孔通过率与空隙率的关系.研究表明,2.36 mm和4.75 mm两个关键筛孔与OGFC混合料空隙率有很好的线性关系.     

交通运输与经济发展关联性分析

随着我国社会经济的快速、深入发展,交通运输对经济发展的作用越来越明显.科学地探讨这种关系,并运用这种关系去能动地指导我们的规划、计划和发展,具有十分重要的意义.本文通过主成分分析法和回归分析法,建立交通运输与经济发展的数学模型,揭示交通运输与经济发展之间相互制约、相互促进的内在密切联系.     

城市道路交通量短时预测的GSVMR模型

在分析现有城市道路交通量短时预测方法缺陷的基础上,针对目前广泛采用的基于经验风险最小化的BP网络易于陷入局部最优解等缺点,结合遗传算法容易寻找全局最优解与支持向量机回归法具有结构风险最小化的特点,提出了将两种算法相结合的GSVMR预测模型,该模型同时具有结构风险最小和容易寻找最优解的双重特性,并对某城市四车道主干道路8：00,8：45的交通量进行了预测,结果表明用该模型进行城市道路交通量短时预测所得结果误差较小,依此验证了用GSVMR模型进行城市道路交通量短时预测的有效性。     

基于驾驶员行为的强制控速设施安装间距研究

在合理分析驾驶员通过强制控速设施行为的基础上,制定了相应的车速检测方案,并进行了实地车速检测。在分析车速与车辆到强制控速设施距离的散点图的基础上,建立了分段的多项式回归模型,然后利用最小二乘法回归计算,得出了相应模型的参数估计值,将模型参数代入回归模型即可获得车速预测函数。最后,进行了回归结果分析,分析结果表明分段多项式回归模型是正确有效的。利用该模型可以有效计算在给定公路限速条件下的最优强制控速设施安装间距。     

Using stop bar detector information to determine turning movement proportions in shared lanes

Turning vehicle volumes at signalized intersections are critical inputs for various transportation studies such as level of service, signal timing, and traffic safety analysis. There are various types of detectors installed at signalized intersections for control and operation. These detectors have the potential of producing volume estimates. However, it is quite a challenge to use such detectors for conducting turning movement counts in shared lanes. The purpose of this paper was to provide three methods to estimate turning movement proportions in shared lanes. These methods are characterized as flow characteristics (FC), volume and queue (VQ) length, and network equilibrium (NE). FC and VQ methods are based on the geometry of an intersection and behavior of drivers. The NE method does not depend on these factors and is purely based on detector counts from the study intersection and the downstream intersection. These methods were tested using regression and genetic programming (GP). It was found that the hourly average error ranged between 4 and 27% using linear regression and 1 to 15% using GP. A general conclusion was that the proposed methods have the potential of being applied to locations where appropriate detectors are installed for obtaining the required data. Copyright © 2016 John Wiley & Sons, Ltd.     

Data Review of Transportation Infrastructure Public–Private Partnership: A Meta-Analysis

The limited understanding of public–private partnerships (P3s) for transportation infrastructure finance has been generally attributed to a lack of data. The more fundamental question of how P3 data are utilized in the literature is more relevant and critical, but remains unclear. This study investigates this question by examining the linkages between research objectives and data characteristics through a meta-analysis of infrastructure P3 studies using multinomial regressions. It analyzes 95 empirical studies that adopt actual data, selected from a P3 research database that includes over 345 studies and are classified into five categories including performance, contract, risk, value for money, and institutional factors. Results show that the case studies are less frequently utilized to understand P3s' institutional issues compared to those that focus on P3s' performance or VFM. Survey data are more frequently used to study P3 contracts rather than issues related to P3 project risk. We highlight the need for policy-makers to require continuing disclosure of P3 performance for validating the effectiveness of the procurement model and to improve the practice.     

Evaluating safety effectiveness of roundabouts in Abu Dhabi

With more than 460 roundabouts located in Abu Dhabi, the capital city of the United Arab Emirates, it is imperative to evaluate the safety benefits provided by those roundabouts. In this study, two approaches were used to evaluate those safety benefits. The first approach is by measuring the 85th percentile operating speeds at a sample of 18 roundabouts in Abu Dhabi to determine whether the measured operating speeds conform to what is recommended by design guides. The second approach is by using a questionnaire to measure how drivers in Abu Dhabi perceive safety when driving at roundabouts and to measure their level of knowledge regarding the rules pertaining to driving at roundabouts. The study found that operating speeds at Abu Dhabi roundabouts typically exceed those recommended by design guides. The study also found that only 4.1% of the drivers interviewed demonstrated a comprehensive understanding of the rules pertaining to driving at roundabouts. Ordinal regression modeling was used to identify driver groups in need for more awareness of the rules to negotiate roundabouts in Abu Dhabi. The study found that the driver group in most need for more awareness is typically young and middle-age male drivers (below the age of 46?years) with less driving experience in countries where roundabouts are common. The questionnaire also revealed that despite the operational benefits provided by roundabouts (in terms of reduced delay), drivers do not prefer to drive at Abu Dhabi roundabouts, which might be explained by drivers' perception of Abu Dhabi roundabouts being not safe.     

改进SVR及其在铁路客运量预测中的应用

为了提高铁路客运量现有预测方法的预测能力,用训练样本与测试样本间的马氏距离对惩罚因子进行加权,对传统的支持向量回归机(SVR)进行了改进,在此基础上提出了基于改进SVR的铁路客运量时间序列预测方法.以1980~1998年铁路客运量预测为例,对SVR方法和BP人工神经网络(BPANN)方法进行了比较,结果表明,SVR方法能获得更准确的预测结果.     

基于支持向量分类机和回归机的综合评价方法

采用支持向量多值分类机和回归机进行综合评价排序,以提高机器学习方法的综合评价排序能力,并以管理信息系统综合评价为例,与人工神经网络（ANN）方法进行了对比研究.试验结果表明,基于支持向量多值分类机综合评价得分之间的差异比ANN更明显,而且基于支持向量回归机综合评价得分的相对误差明显小于ANN.