基于MATLAB的交通事故发生规律的模糊分布拟合

借助于数学软件MATLAB,运用模糊数学方法,探讨中国交通事故发生及危害的变化规律.首先,作中国交通事故统计数据的散点图,观察发现交通事故的变化同时具有线性增长总趋势和波动性特征,呈现出复杂的非线性组合特点;再对这些数据进行归一化处理,给出交通事故的模糊隶属函数;最后对交通数据进行分布拟合、组合回归处理,得到交通事故非线性模糊分布函数.结果表明,随着时间的变化,中国交通事故的死亡人数呈线性与正态叠加波动上升趋势,受伤人数呈线性与正弦叠加波动上升趋势,发生起数与损失折款都呈双峰正态分布波动趋势,2002年以来交通状况正在好转.得到的结果能客观地反映交通事故的现状分布和精确地刻划中国交通事故发生及危害变化规律,对于交通事故预测、综合治理具有重要的理论意义和应用价值,可供交通事故研究和有关部门制定交通安全管理决策作参考.     

北京市道路交通事故灰色预测模型的构建与应用

利用灰色预测理论,根据北京市1999～2006年道路交通事故的统计资料,建立了北京市道路交通事故发生次数的灰色GM(1,1)预测模型,并根据实际数据的特点,对所建立的模型进行了改进.经证实,改进后的GM(1,1)模型预测精度高,预测结果与实际相符.GM(1,1)模型的预测结果可为北京市道路交通管理部门制定预防交通事故的措施提供重要的数据支持和理论依据,这对加强北京市的道路交通管理、改善北京市的道路交通安全状况将具有一定的现实意义.     

基于BP神经网络的道路交通事故预测

道路交通事故预测是交通研究的一个重要课题.以我国交通安全状况为研究对象,依据我国道路交通事故的特点,利用神经网络具有自学习、自组织、自适应能力特性,运用神经网络的方法及我国多个年度道路交通事故统计数据,建立了道路交通事故神经网络宏观预测模型,预测精度符合道路交通事故预测的要求.     

北京市道路交通事故致死率的分析与预测

根据北京市道路交通事故的统计资料,以交通事故致死率为指标,对北京市目前道路交通安全情况进行评价.并基于灰色预测理论,建立北京市道路交通事故死亡人数的灰色GM(1,1)预测模型.预测结果显示,北京市未来几年道路交通事故死亡人数虽处于下降趋势,但死亡人数仍在千人左右,平均每起事故死亡人数仍处于上升趋势,道路交通事故致死率仍然偏高.道路交通管理部门应加大惩治道路交通违法行为力度,以控制道路交通事故的严重后果.     

A general model of road traffic accidents

A general model that relates road traffic accidents to the number of vehicles involved, and the number of primary causes of such accidents, is presented. The model considers traffic accidents as failures of a road traffic network system to meet social and economic constraints, and therefore as a measure of the unreliability of such a system. The equations apply to accidents in the real time domain as well as to mean values per unit of vehicle exposure time or vehicle distance. They also apply to single vehicles and drivers, groups of drivers and fleets of vehicles, and the entire vehicle and driving population. They can be used for sections of a network or for a whole network.The equations used have a large number of terms, hence bias errors are common in road accident investigations.     

灰色关联分析在中国交通事故综合评价中的应用

利用数学软件M ATLAB制作了一个多元灰色关联分析工具箱,并用该工具箱对1997~2006年中国31个地区交通事故统计数据进行多元灰色关联分析,从18个社会经济因素中找出了对交通事故指标影响关系最为密切的几个社会经济因素,并对中国31个地区10年交通事故危害程度作综合评价.研究表明:各地区"生产总值"及"旅客周转量"等社会因素对交通事故四个指标(发生起数、死亡人数、受伤人数和损失折款)影响最大;多考虑"生产总值"及"旅客周转量"两个指标对31个地区的交通事故危害程度的综合评价,比仅仅考虑四个交通事故指标的综合评价,结果更为合理;1997~2006年中国交通事故最为严重的前3个地区为:广东、浙江、福建.研究结果可供交通管理部门参考.     

On the derivation of the velocity and fundamental traffic flow diagram from the modelling of the vehicle–driver behaviors

This paper deals with derivation of the fundamental diagram by modelling the individual driver behavior that adjusts the velocity to the density of vehicles in order to respect the braking distance. A parameter is properly introduced to model the quality of the driver–vehicle subsystem referred to the environmental conditions. Subsequently, it is shown how to use this result in order to model traffic flows by the macroscopic representation and by the kinetic theory.     

基于判别分析的骑推行交通事故鉴定研究

道路交通事故鉴定对于交通事故责任认定和法庭举证具有重要作用.采集某市近三年125起交通事故案例数据,建立机动车、自行车、人体总计73个调查变量,运用SPSS软件开展骑推行事故鉴定研究.通过相关性分析发现车座、机动车类型等6个变量与鉴定结论相关.开展线性判别分析研究,结果表明直接对相关变量进行判别分析可以快速获得最佳判别效果.当函数选入的自变量为车座时,交叉验证准确率最高可达72.8%,说明采用数据挖掘的方法来鉴别交通事故中的行为方式具有一定可行性.     

道路行车系统中人的可靠性研究

随着人-车-路-环境系统复杂性的增加和交通事故数量的上升,新型交通设备对其使用者的素质与适应性提出了全面的要求,要提高道路行车系统的整体可靠性,必须对人的可靠性进行分析研究.从人的生理、心理、教育训练、驾车技能等因素对驾驶员可靠性的相关影响进行了探讨.分析了影响驾驶员的可靠性因素,研究了驾驶员的可靠性模型,提出建立有助于驾驶员可靠性分析研究的综合数据库,最后从理论上分析总结了提高驾驶员可靠性的措施从而得出提高道路行车安全的有效途径.     

基于GA-ELM的水上交通事故严重程度影响因素识别研究

水上交通事故严重程度影响因素的识别对从根本上减少严重事故件数、降低事故危害和损失具有重要意义。在历史事故报告的基础上,构建并量化事故影响因素集,提出以极限学习机(ELM)为一般事故、严重事故的二分类器,以遗传算法(GA)为因素搜索算法的GA-ELM因素识别模型。对发生在我国水域的737件水上交通事故进行实证研究,并与以支持向量机(SVM)为分类器的GA-SVM模型进行对比分析。结果表明,GA-ELM模型识别出时段、人为致因、环境致因等9个事故严重程度影响因素,较GA-SVM模型结果更为精简,且分类精度较不做因素识别时分别提高8.2%、7.1%。此外,GA-ELM大大缩短运算时间。由此可见,GA-ELM可为水上交通事故严重程度影响因素识别提供一个较好的方法。     

The design of a Tort law to control accidents

In this paper we analyse a two-stage game involving the government and n agents who engage in a single activity (driving). The government establishes the legal policy setting and the agents proceed to play a non-cooperative game of incomplete information with a risk of accident in which their behavioral strategy is their level of care. We examine the Nash-equilibrium conditions for single-activity accidents between heterogeneous agents, ‘good’ drivers or ‘bad’ drivers allowing a variable damage function and a liability rule defined on the cube. The relative desirability for society of alternative equilibria and the conditions under which they can obtain are discussed. The constraints which circumscribe the ability of the government to induce an equilibrium involving careful driving are demonstrated. It transpires that when the proportion of good drivers increases, it becomes more difficult to sustain a careful equilibrium whereas an equilibrium of reckless behavior becomes easier to sustain. Various extensions of the models are also presented.     

An algorithm to assign pedestrian groups dispersing at public gatherings based on pedestrian-traffic modelling

The minimization of weighted pedestrian and vehicle times for pedestrians dispersing at public gatherings is considered. An algorithm to solve this problem and to assign each pedestrian group to a unique selected path is developed. This algorithm considers general network where its arc values are dependent on pedestrian/traffic modelling and its nodes (intersections) have penalties.The algorithm is followed by a case study which exhibits possible pedestrian/traffic models. These experimental models yield some deterministic functional dependency between pedestrians' speed and sidewalk concentration, pedestrians' street concentration, and sidewalk concentration, and vehicles' speed and pedestrians' street concentration. If such and realistic models serve as an input to the algorithm it might contribute to reduce pedestrian/traffic conflicts and interference and hence to increase the overall safety of these two travel modes.     

生产安全事故宏观发展态势的计量经济学研究

为了研究工矿商贸就业人员10万人生产安全事故死亡率时间序列变化特征,基于我国行业生产安全事故死亡人数及第二、三产业就业人员数量等2方面年度统计数据,通过研究事故死亡率时间序列的自回归移动平均过程,论文构建了事故死亡率时间序列的分阶段自回归移动平均模型.研究表明:工矿商贸行业10万人事故死亡率变化趋势具有明显的分阶段波动特征,事故死亡率序列均为趋势平稳过程;序列当期观测值与滞后1期观测值具有显著的自相关性;各阶段事故率自回归移动平均模型结构不尽相同;特征描述模型为正确把握我国安全生产状况及趋势提供理论依据.     

Kink–antikink density wave of an extended car-following model in a cooperative driving system

We propose an extended optimal velocity model applicable to cooperative driving control system by considering the headway of arbitrary number of cars that precede and the relative velocity. The stability condition of the extended model is obtained by using the linear stability theory. The modified Korteweg-de Vries (mKdV) equation is derived to describe the traffic behavior near the critical point by applying the nonlinear analysis. Thus the traffic jams can be described by the kink–antikink density wave which is the solution of the mKdV equation. The simulation results confirm the analytical results and show that the traffic jams are suppressed more efficiently with considering not only the headway of more vehicles ahead but also the relative velocity.     

A new lattice model of traffic flow with the consideration of individual difference of anticipation driving behavior

In this paper, we propose a new lattice model of traffic flow with the consideration of individual difference of anticipation driving behavior. The linear stability condition and the mKdV equation are derived from linear stability analysis and nonlinear analysis, respectively. Furthermore, numerical simulation shows that the anticipation driving behavior can increase the cell number of low density, which means that more cars can run freely and traffic congestion can be suppressed efficiently by taking the anticipation driving behavior into account in lattice model. Moreover, with the coefficient of the anticipation driving behavior increasing, the low density region turns wide corresponding to individual difference of anticipation driving behavior.     

缆机起重作业驾驶员的操作行为可靠性研究

缆机是大坝浇筑过程中的主要运输设备之一,缆机驾驶员操作水平直接关系缆机起重作业安全.为有效提高缆机起重作业驾驶员操作行为可靠性,预防起重作业安全事故,针对缆机起重作业周期性多维运动的特点,分解驾驶员操作流程;剖析操作行为失误,建立驾驶员行为模式;从时间角度量化操作流程,计算缆机驾驶员操作行为响应失误概率;构建操作行为可靠性分析模型,定量表达操作行为可靠性,分析人因失误对缆机起重作业安全事故的影响.结果表明,吊具与吊罐对接、混凝土卸料等作业过程的缆机驾驶员操作行为可靠性低,需加强安全培训,严格考核制度,从根本上提高缆机驾驶员操作行为的可靠性,以期提升缆机生产效率,实现预防安全事故的目标.     

Driving while intoxicated (DWI) convictions and job-related flying performance—a study of commercial air safety

This paper is a report on the relationship between driving while intoxicated (DWI) convictions and job-related flying performance. The data, obtained by the Federal Aviation Administration, focuses on commercial air accidents and incidents in the years 1986–1992. Pilot-error accidents were modelled using logistic regression with age, experience (total flying hours), risk exposure (recent flying hours) and employer (major/non-major) as the independent variables. This study provides empirical evidence that job-related flying performance differs significantly between airline pilots with and without DWI convictions. It also suggests that passengers on a commercial flight flown by a pilot with DWI(s) may incur an increased accident risk. DWI background checks may be an effective method for improving commercial air safety if used with appropriate follow-up actions.     

安全生产事故的时间特征分析

基于我国安全生产事故快报数据,采用季节指数法对近年安全生产事故现状给出时间特征分析,确定出安全生产事故高发时间段.同时,给出总体的和死亡1人、3人以及4人的安全生产事故每年在每个时段平均事故总起数的概率分布模型.分析方法和结果有助于国家安全生产事故预防、控制和应对的监管工作.     

极端损失条件下分组群体的主观概率累积偏差仿真研究

为了研究个体死亡这种极端损失条件下分组群体的主观概率累积偏差的形成与演化规律,构建不同分组方式下的群体主观概率元胞自动机模型并运用Netlogo仿真平台进行实验。在对仿真结果进行图形分析的基础上,进一步对实验数据进行了回归分析和案例验证分析。结果表明:分组群体的主观概率演化一般会经历从急速变化到趋于稳定的过程;分组作业方式能够提高安全冗余,但不利于个体事故经验积累;个体死亡等极端损失会导致群体的主观概率低估和分化,但分组措施能够降低此作用;客观事故概率的减少会增大群体内部主观概率的差异,但分组措施能够降低此作用;事故率与死亡率之间存在互补效应,分组措施对其存在双向调节作用;群体人数增加有助于减少主观概率偏差和差异。     

Cost-benefit analysis of one-server n-unit adaptive system with slow switch

This paper considers a shared parallel system consisting of n-units supported by single service facility to carry out both installation and repair of a unit. Initially, all the n units share the total load equally and when one or more units fail, they go for repair while the other surviving units share the entire load equally till the failed units are ready for operation after installation. The installation time (switchover time) of a repaired unit is assumed to be non-negligible and random. The system will be down when all the units are non-operative , Assuming that the failure rates are different when the units function under varying loads, the system characteristics, namely, (1) the expected up-time of the system during (0, t], (2) the expected repair time of the units which failed due to varying failure rates during (0, t] and (3) the expected time spent by the units in the installation state during the period (0, t], are obtained by identifying the system at suitable regeneration epochs. The repair time and the switchover time of the units are arbitrarily distributed. The failure rate of unit is assumed to be constant. It depends on the number of surviving units at any instant. The cost-benefit analysis is also carried out using these system characteristics