双车道等级公路路侧事故预测模型研究

借鉴国内外事故预测模型(也称安全性能函数SPF)研究领域的最新成果,收集了31条双车道公路(总里程约740 km)的事故、交通量和线形数据,采用了泊松、负二项、零堆积泊松和零堆积负二项四种统计概率分布,从路侧事故数、路侧事故死亡人数、路侧事故受伤人数三方面对路侧事故规律进行了研究,通过弹性分析从量化的角度给出了道路几何条件、交通量水平与构成等指标与路侧安全的关系。     

Modèles linéaires généralisés appliqués à l'étude des nombres d'accidents sur des sites routiers: Le modèle de Poisson et ses extensions

Models that link accident frequencies at road sites with traffic volumes and road characteristics provide reference values which are of value when carrying out diagnostic studies and deciding on preventive measures. Such models can be developed using generalized linear modelling techniques, which are able to take account of the Poisson properties of accident frequencies at road sites. The Poisson models obtained for road sites are frequently everdispersed, so extended forms of the Poisson model (quasi-Poisson, negative binomial a high percentage of the studied sites, in which case zero-inflated models also have to be considered. This paper presents an overview of these methods which are illustrated by modelling the relationship between accident frequency site characteristics and traffic volume on a large sample of bends.     

零截尾负二项模型在交叉口事故预测中的应用

事故预测模型是广泛采用的交通安全定量分析方法，但往往要求具有完备的道路、交通和事故数据。然而，基础数据相对不健全是包括中国在内的发展中国家交通安全管理面临的主要问题之一，例如仅有发生事故路段或者交叉口的相关属性特征（即零截尾数据）。为此，为确保基础数据不全的情况下交叉口事故预测的准确性，提出了基于零截尾的广义负二项回归模型；采集了246个非信号控制交叉口的交通与事故数据，采用传统负二项模型和新提出的零截尾负二项模型对全数据和零截尾数据分别进行对比分析。结果表明：在针对截尾数据的分析中，零截尾负二项模型明显优于传统负二项模型，并且零截尾负二项模型的参数估计值与基于全数据的负二项基准模型的估计值非常接近；在所有模型中，交叉口的主路交通量和支路交通量与交叉口的安全性之间存在较大的正关联。此外，同等条件下，十字形交叉口的事故数量高于T形交叉口的事故数量；利用传统负二项分布模型分析截尾数据得到的事故预测模型与使用全数据的基准模型有显著差异，其结果不可靠；采用零截尾负二项分布模型的参数结果与基准模型基本一致，截尾模型的置信区间包含基准模型相应的参数估计值。当受条件所限无法获取全部数据时，可以考虑使用零截尾负二项模型进行安全分析。     

基于有限混合零截尾事故预测模型的事故多发段判别方法

为了准确判别事故多发段,有针对性地提出安全应对措施以提升道路交通的安全水平,针对零值缺失交通事故数据并考虑其异质性特点,在单零截尾负二项(ZTNB)模型的基础上建立有限混合零截尾事故预测模型(FMZTNB)。应用R软件对单零截尾负二项模型中的参数进行估计,采用马尔科夫链蒙特卡洛算法(MCMC)对FMZTNB预测模型参数进行求解,并采用Gelman-Rubin收敛统计量对抽样结果进行检查。选择事故风险水平分别为低、中和高的9个路段,分别用2种模型对交通事故次数进行预测。综合观测到的事故次数和相应的事故预测模型结果,采用经验贝叶斯方法对事故相对多发段进行判别。最后采用事故次数一致性检验、判别点段一致性检验和排序一致性检验3种检验方式对判别结果对比分析。结果表明:基于事故率的事故相对多发段判别方法存在较大的不一致性,基于零截尾负二项预测模型的路段事故相对多发判别结果明显优于基于传统负二项预测模型的结果。整体上,基于有限混合零截尾事故预测模型的事故相对多发路段的判别结果高于基于单零截尾负二项分布模型的判别结果。     

基于贝叶斯空间相关模型的城市快速路安全影响因素研究

快速路是城市交通的主骨架,其运行安全态势直接影响整体通行效率的发挥。利用线圈检测器数据和由事故视频监控系统采集的2011~2013年共50069起事故数据,应用贝叶斯空间相关模型,对上海市南北高架、延安路高架、内环和中环4条快速路进行建模分析。为了反映路段间的潜在空间联系,在负二项模型的基础上,加入空间随机影响项建立相邻空间相关模型;进一步改进了该空间随机项的协方差矩阵,建立距离空间相关模型。通过贝叶斯回归方法比较3个模型的回归效果。结果显示,距离空间相关模型具有最优的拟合效果,其方差信息标准(DIC)较负二项模型降低6.04%。由此证明,事故在道路中的分布存在空间关联,且其相关性受距离影响。在道路几何特性影响因素中,路段前后1 km 内匝道数越多,每公里转角越大,事故风险越大;入口-出口型交织路段事故发生几率较高。日均交通量与事故风险显著正相关。不同快速路对比分析表明内环事故风险偏高,中环事故风险较低。      

基于负二项分布的高速公路交通事故影响因素分析

为分析高速公路交通事故的影响因素,构建基于负二项分布的事故分析模型,探究事故数与交通特性、公路线形及路面性能间关系.鉴于传统固定参数模型难以刻画各因素对事故风险影响的异质性,引入了随机参数建模方法.结果表明:相比于固定参数负二项模型,构建的随机参数负二项模型有更好的拟合优度,且能更合理地反映各因素对事故的作用效果;将随...     

考虑多模式失效概率的长下坡路段重型卡车事故预测模型

为挖掘多模式失效概率与长下坡路段重型卡车事故之间的关系，建立了重型卡车在长下坡路段的多模式失效概率与车辆事故之间的关系模型。并针对重型卡车在长下坡路段可能的失效模式，如侧滑、侧翻、视距不足、制动失效，在此基础上建立了多模式失效概率预测模型；通过蒙特卡罗法模拟并求解单模式失效的概率，宽界限法求解失效系统的多模式失效概率；将多模式失效概率作为解释变量与其他道路因素结合，分别建立泊松模型、随机效应泊松模型、随机参数泊松模型，将多模式失效概率与重型卡车事故建立函数关系；对比3种模型的拟合优度指标，优选出最优事故预测模型，用来挖掘重型卡车事故与多模式失效概率之间的关系。以华盛顿州71段长下坡10年的重型卡车事故数据及道路设计数据进行方法验证。结果表明:随机参数泊松模型与随机效应泊松模型的拟合优度相差较小，二者均优于泊松模型；当考虑多模式失效概率时，平曲线半径、纵坡坡度、超高对重型卡车事故的影响均不显著，即三者的影响被削弱，尤其是平曲线半径和超高，多模式失效概率的弹性(0.239)远大于二者的弹性(平曲线半径和超高的弹性分别仅为0.097和0.002)；重型卡车的事故与多模式失效概率近似线性关系，且截距不为0。即多模式失效概率可用于道路安全分析的表征指标，但与事故概率不等价。      

Predicting downgrade crash frequency with the random-parameters negative binomial model: Insights into the impacts of geometric variables on downgrade crashes in Wyoming

Road deaths, injuries and property damage place a huge burden on the economy of most nations. Wyoming has a high crash rate on mountain passes. The crash rates observed in the state is as a result of many factors mainly related to the challenging mountainous terrain in the state, which places extra burden on drivers in terms of requiring higher levels of alertness and driving skill. This study was conducted to investigate factors leading to crashes on Wyoming downgrades, with a focus on geometric variables. Traditionally, crash frequency analysis is conducted using count models such as Poisson or negative binomial models. However, factors that affect crash frequency are known to vary across observations. The use of a methodology that fails to take into account heterogeneity in observed and unobserved effects relating to roadway characteristics can lead to biased and inconsistent estimates. Inferences made from such parameter estimates may be misleading. This study employed the random-parameters negative binomial regression models to evaluate the impact of geometric variables on crash frequency. Five separate models were estimated for total, fatal/injury, property damage only (PDO), truck, and non-truck crash frequencies. Several geometric and traffic variables were found to influence the frequency of crashes on downgrades. These included segment length, vertical grade, shoulder width, lane width, presence of downgrade warning sign, vertical curve length, presence of a passing lane, percentage of trucks, number of lanes and AADT. The results suggest that segment length, lane width, presence of a passing lane, presence of a downgrade warning sign, vertical grade, and percentage of trucks are best modeled as random parameters. The findings of this study will provide transportation agencies with a better understanding of the impact of geometric variables on downgrade crashes.     

高速公路交通事故非线性负二项预测模型

以京港澳高速公路（G4）粤境北段3年发生的1 354起交通事故为研究对象,将基础数据根据路段长度一致、曲线半径一致和坡度一致划分路段单元对基础数据进行处理,从道路线形和环境条件2个方面选取13个自变量,分别采用负二项（Negative Binomial,NB）回归模型和非线性负二项（Nonlinear Negative Binomial,NNB）回归模型建立交通事故起数预测模型,根据模型的拟合优度和预测准确性对比分析负二项回归和非线性负二项回归模型的优劣,并找出影响交通事故起数的显著自变量,分析显著自变量对交通事故起数的影响程度。研究结果表明：无论采用上述何种路段划分方法,非线性负二项回归模型构建的交通事故起数预测模型均优于负二项回归模型;采用坡度一致划分方法明显优于路段长度一致和曲线半径一致划分方法,更适合应用于山区高速公路交通事故数预测研究;从显著变量相关性来看,路段长度、相邻路段坡度变化值、弯坡组合、曲率、是否存在隧道路段以及是否为易结冰和起雾路段均是非线性模型的显著影响因素。     

高速公路交通事故模糊逻辑预测模型改进研究

以AADT、路段长度、车道数、大型车比例和地形条件作为模型输入变量,以每公里事故数作为模型输出变量,结合辽宁省高速公路数据构建基本的交通事故模糊逻辑预测模型.考虑到模糊集合结构和模糊控制规则对预测结果可能产生的影响,提出调整模糊集合和融入先验知识构建规则库的模型改进方法.以粤赣高速和开阳高速为案例,分析了基本模型与改进模型的可移植性.最后,应用同样的数据构建了负二项分布事故预测模型,并与模糊逻辑预测模型进行了对比分析.研究结果表明,纵向比较,模糊集合细化一定程度提高模型预测精度,细分模型相较于基本模型,总体平均相对误差减少8.3%,模型优度提高0.357;横向比较,融入先验知识构建模糊规则库能一定程度提高模型预测精度,基本先验模型相较于基本模型,总体平均相对误差减少1.9%,模型优度提高0.164.融入先验知识后模型的可移植性增强,平均预测精度高于基本模型,相对误差大于0.5的样本数减少3.8%,总体误差减少3.4%,总体平均相对误差减少4.1%,模型优度提高0.385;但细化集合的模型可移植性较低,与粗分和基本模型相比各个指标值均不同程度变差;而模糊逻辑事故预测模型与负二项分布事故预测模型在预测精度和可移植性方面均无显著差异.      

交通暴露与土地利用模式对社区COVID-19传播风险的影响

为揭示交通暴露和土地利用层面各因素对新型冠状病毒肺炎（COVID-19）疫情在社区尺度传播风险的影响，以A市315个小区的1 947例COVID-19确诊数据为研究样本，利用ArcGIS平台的地理编码、核密度分析、空间统计和网络分析等方法，基于道路网络、公交网络、城市POI和中国GDP/人口空间分布的公里网格数据，获取有病例小区500 m缓冲区范围内表征交通暴露和土地利用的14项具体指标。在此基础上，采用小区的COVID-19确诊人数作为解释变量，同时考虑交通暴露层面变量（路网密度、设施邻近度）和土地利用层面变量（混合度、使用强度），运用经典全局泊松回归和变系数地理加权泊松回归（GWPR）2种方法建立模型并进行实证对比分析。研究结果表明：考虑空间异质性的GWPR模型具有更高拟合优度和解释度；道路密度、公交线网密度、CBD邻近度、建筑密度、人口密度和土地价值与小区COVID-19传播风险呈显著正相关；出入口邻近度、绿地公园邻近度和土地利用混合度变量则在GWR模型中表现出随空间位置的改变呈现显著正负2种影响效果；人口密度、土地价值、绿地公园邻近度和土地利用混合度对小区COVID-19传播风险的影响程度要高于其他变量。因此，城市空间要素不仅会影响非传染性的病发风险，同时也与传染性疾病的传播风险显著相关，所得结果可为通过土地利用、城市交通规划等手段降低流行病发生的潜在风险提供参考。     

Modeling two-way stop-controlled intersection crashes with zero-inflated models on Louisiana rural two-lane highways

Intersection safety continues to be a crucial issue throughout the United States. In 2016, 27% of the 37,461 traffic fatalities on U.S. roadways occurred at or near intersections. Nearly 70% of intersection-related fatalities occurred at unsignalized intersections. At such intersections, vehicles stopping or slowing to turn create speed differentials between vehicles traveling in the same direction. This is particularly problematic on two-lane highways. Research was performed to analyze safety performance for intersections on rural, two-lane roadways, with stop control on the minor roadway. Roadway, traffic, and crash data were collected from 4148 stop-controlled intersections of all 64 Parishes (counties) statewide in Louisiana, for the period of 2013 to 2017. Four count approaches, Poisson, Negative Binomial (NB), Zero-inflated Poisson (ZIP) and Zero-inflated Negative Binomial (ZINB) were used to model the number of intersection crashes for different severity levels. The results indicate that ZIP models provide a better fit than all other models. In addition to traffic volume, larger curve radii of major and minor roads and wider lane widths of major roads led to significantly smaller crash occurrences. However, higher speed limits of major roads led to significantly greater crash occurrences. Four-leg stop-controlled intersections have 35% greater total crashes, 49% greater fatal and injury crashes, and 25% greater property damage only (PDO) crashes, relative to three-leg intersections.     

不同优化类型组合预测模型在沙漠软土路基的应用

某沙漠重载铁路工程，地势低洼路段雨季易积水，加上原有高地下水位的长期浸泡和植物根系的有机分解积累，形成了淤泥质或泥炭质软土地基；运用理论分析、现场监测等方法，从权重分配合理性和子模型组合结构两方面探讨组合预测模型的精度优化效果，开展沙漠淤泥质软土路基沉降预测研究。结果表明：双曲线法、三点法、指数曲线法、泊松曲线法、BP神经网络5种预测模型均能达到较高水平的拟合程度；变权重组合预测模型、引入鲸鱼优化算法的自适应权重组合预测模型、滚动动态组合预测模型对于预测精度、效果的提升较小；引入BP神经网络的误差补偿组合预测模型，极大程度地降低了人为建立子预测模型组合结构所产生的精度影响，在沙漠淤泥质软土路基中具有更优的预测精度及效果。     

Evaluating the correlation between risky riding behaviour and self-reported crashes in India: Minimizing unobserved heterogeneity

Riding Behaviour is found to be the main cause of Powered Two-Wheeler (PTW) crashes in more than90% of the crash events. The high percentage of PTW crashes resulting in fatalities has sought a serious need for research to examine risky riding Behaviour. A widely used instrument for measuring the self-reported riding Behaviour of PTW riders is the Motorcycle Rider Behaviour Questionnaire (MRBQ). In this study, exploratory factor analysis of the MRBQ revealed a four-factor solution viz., traffic errors, control errors, speed violations, and stunts. Despite the popularity of MRBQ, it is capable of covering only a small fraction of the large number of elements that affect the riding Behaviour. Many other elements remain overlooked in the analysis, resulting in unobserved heterogeneity. Therefore, the present study uses a random parameter negative binomial (RPNB) model to minimize the effect of unobserved heterogeneity. It was inferred from the RPNB model that variables like gender, control error, and speed violation have a randomly distributed regression coefficient. Further, it is found that traffic errors are the most significant predictor of crash risk. Additionally, results depict that male riders are positively associated with crashes, and they are more likely to involve in crashes as compared to female riders. The finding of this paper will be valuable for policymakers and decision-makers to improve the rider training program, licensing system, and design road safety campaigns.     

Pedestrian accessibility and safety around bus stops in Delhi

Safety for public transport (PT) users is least during the access and egress trips. Previous studies have established that improvement in pedestrian safety will improve the safety of PT users as well. We studied the accessibility of pedestrian infrastructure around ~360 sampled bus stops in Delhi by conducting physical audits. 15 indicators in the audit checklist were meaningfully reduced to five factors through Principal Components Analysis. We developed Poisson regression models (with their geographically-weighted counterparts) to assess the association between these five factors for each bus stop with the number of pedestrian fatalities around that stop. Two models were developed— a) for fatalities where the impacting vehicle was known, and b) for fatalities where the impacting vehicle was unknown (hit-and-run cases). For both the models, geographically-weighted Poisson regression (GWPR) performed better than their global Poisson counterparts. Overall improved access was seen to be positively associated with less pedestrian fatalities. Further, we established that the nature of hit-and-run cases differ from those where the impacting vehicle is known, through— a) difference in the effect of the exposure variable, b) different factors being significant in the respective models, especially in the GWPR. The novelty of this study is that we modelled the relationship of pedestrian fatalities around PT stops with factors related to the pedestrian access to these stops. Through the application of GWPR, we found that different types of pedestrian fatalities are related to different aspects of access. We also identified bus stops with higher risk of pedestrian fatalities. Based on this, the methodology presented in this study is useful in guiding city authorities to identify and prioritise a) specific access-related factors which require improvement, and b) bus stops which require improvement in their pedestrian-access infrastructure. These analyses can be extended to study pedestrian safety around PT stops in any city.     

泰州长江大桥钢塔疲劳计算随机车辆荷载模拟

通过对江阴长江大桥实际行驶车辆的概率统计分析,采用分段多项式曲线拟合方法得到了货车和客车的随机车重概率近似表达式。通过泊松方法建立车辆随机间距的数学模型,将采用数学模型的计算结果与实际观测数据进行了对比分析,表明了建立的数学模型的有效性。基于以上的数学模型和随机过程理论,建立了泰州长江公路大桥钢塔疲劳计算随机车辆荷载模型,以便能够更好的重现公路桥梁实际车辆的统计特性。     

基于线形与交通状态的山区高速公路追尾事故预测

为了进行山区高速公路追尾事故预测并识别追尾事故突出诱导因素,在对两车追尾事故进行类别划分并确定出典型两车追尾事故的基础上,分析了典型两车追尾事故的事故率与线形指标、车速差、大型车混入率、交通量等单一因素间的相关关系。鉴于单一因素与追尾事故率间的关系不能准确描述追尾事故发生规律的缺陷,建立了线形与交通状态组合条件下的追尾事故次数负二项分布预测模型,并给出了模型变量弹性系数计算方法,用以确定追尾事故的突出诱导因素。研究结果表明：基于线形与交通状态的追尾事故负二项分布预测模型能够对追尾事故进行准确预测,利用弹性系数计算方法确定出车速差、年平均日交通量（AADT）以及竖曲线半径为典型两车追尾事故的突出诱导因素。     

Prediction of wheel wear in urban railway transport: comparison of existing models

One of the most important maintenance costs in tramway transport comes from wear of wheel profiles. In the highly competitive railway market, the prediction of wear is then a major concern of the constructors. In this article, we present and compare four models well adapted to tramway conditions, involving contacts on the rolling tread and on the flange with very different sliding and pressure conditions. Moreover, all models can be implemented from the natural outputs of the railway simulation packages classically used in industry for the dynamics design of the vehicles. The first one, proposed by Jendel, is based on the well-known Archard's wear model. Enblom continues Jendel's approach by taking into account the contribution of wheel deformation on the sliding velocity. The last two models, developed by Zobory, and Pearce and Sherratt, determine the wear from the energy dissipation in the contact area. The models are first compared on a theoretical basis and, for that purpose, are rewritten in a common form. Two cases are distinguished: mild wear as arising on the rolling tread and severe wear as arising on the flange. The models are also compared in the practical case of an urban transport vehicle running on circular tracks with different curve radii. Although the models show equivalent trends according to the theoretical study, important discrepancies appear between estimated wear depths. All models are actually dependent on experimental coefficients and it is likely that they were estimated in different conditions. On the other hand, a reasonable agreement can be found in some particular conditions. As an example, Zobory's, Enblom's and Jendel's models are very close to each other in severe wear conditions. This work shows that a general and reliable model could probably be developed from all positive aspects of the existing ones.     

Selfexcited Vibrations of the Tire

In this paper a mechanical work of external forces and torques, acting on the tire has been considered. A theory has been developed for the prediction of necessary conditions for selfexciting vibrations. The theory establishes, that the mechanical work of external forces and torques must be positive, or, what is equivalent, that the tire has to transmit energy from external environment to vehicle. The work of forces and torques has been considered as a convolution of functions, which next has been submitted to Laplace transformation. The condition of selfexcitation of vibrations, obtained from above in the frequency domain, establishes, that the imaginary part of all eigenvalues of frequency response matrix must be negative. As an example, the ranges of selfexcited vibrations of the tire have been calculated. The tire has been treated as a stretched circular string. Four different models have been considered. Three of them have been massles (kinematic), namely Pacejka's model, Von Schlippe's model and single-point model. As fourth model the dynamic Pacejka's model has been considered. The influences of longitudinal deflections on the shimmy tendency have been studied for all models

Frequency response characteristics have been numerically computed and compared for all models. A critical speed and a critical reduced frequency of first and second kind have been defined. Admissible, negative, imaginary parts of the eigenvalues of frequency response matrix have been established. These admissible values of imaginary parts of eigenvalues assure, that the energy absorbing does not excite shimmy vibrations.     

车车通信环境下考虑交通拥堵状态的碰撞时间混合分布建模研究

碰撞时间（TTC）是评价车车碰撞风险的有效指标，然而该指标分布规律受到交通状态影响。为研究车车（V2V）通信环境下不同交通状态的TTC分布规律，通过构建基于LTE-V技术的车车通信环境，开展实车实验获取4种典型城市道路中的驾驶数据。考虑加速度和航向角建立动态冲突辨识模型，计算车辆以任意角度接近时的TTC值；针对TTC值的结果出现多峰值现象，将交通流分为“拥堵、缓行、畅通”这3种状态，构建了考虑交通流状态的高斯混合模型以描述不同交通状态下的TTC分布规律，并采用最大期望（EM）算法进行参数求解。将所建高斯混合模型与负指数分布、对数正态分布、负指数/对数正态混合分布这3种传统的TTC分布模型进行对比，采用校正决定系数R2评价模型的拟合优度，并通过K-S检验验证模型的有效性。在此基础上，将所建高斯混合模型应用于非车车通信条件下不同交通状态的TTC分布拟合描述，进一步验证模型的适用性。结果表明:车车通信环境下“拥堵、缓行、畅通”这3种交通状态下的高斯分布均值逐渐增大，所处交通场景的碰撞风险依次降低；考虑交通状态的TTC高斯混合模型拟合优度为0.950 5，相较于其他TTC混合分布模型，拟合优度提升了0.057 5。