一种基于收费数据的断面流量估算方法

高速公路的流量是反映高速公路交通状态的一个重要参数.目前,针对高速公路断面流量的采集主要依靠人工方式或者通过安装一些车流检测设备来获得,成本较高.结合高速公路收费系统的特点,通过对收费数据的统计分析,提出一种基于收费数据估算高速公路断面流量的方法.首先,对收费数据进行初步的处理,从中计算出行驶时间;然后,根据收费数据中的车辆信息统计出各个OD的交通流量;最后,结合各个收费站出入口的位置信息、路段行驶速度信息以及OD流量来估算断面的交通流量.     

基于双层规划的高速公路收费费率优化模型

在全国联网收费的背景下,从动态收费的角度考虑,建立了双层规划模型,上层规划中将路网管理者作为领导者,以高速公路收费效益最大化为目标函数,同时考虑道路运营管理方的合理收益和养护成本支出情况,下层规划则以用户出行效用最大化为目标,充分考虑了道路使用者的道路选择差异性及道路拥堵对交通分布的影响,建立随机用户均衡模型.最后结合某地区AB地高速公路实际情况进行分析,采用了遗传模拟退火算法验证了模型的实用性,并与其他的算法对比,验证了算法的有效性。研究表明:优化模型可以有效提高高速公路的收费效益和用户的出行效用,可以分散高峰时的交通压力,提升高速公路的通行效率.     

基于CVaR的次优拥挤收费随机多目标模型

次优拥挤收费问题一般要考虑不同决策者的不同利益,因此,有必要考虑多个收费策略建立多目标模型来均衡不同决策者的利益.由于决策者常在信息不确定的情况下做决策,在出行需求不确定的条件下,为了确定次优拥挤收费的方案,建立了基于条件风险价值的随机多目标双层规划模型,上层规划的目标函数考虑了系统总阻抗和社会公平性,下层规划是UE用户均衡配流问题.利用基于随机模拟的遗传算法对模型进行求解,并通过数值算例对模型和算法进行分析,验证了模型的有效性.     

基于CVaR的次优拥挤收费随机多目标模型北大核心

次优拥挤收费问题一般要考虑不同决策者的不同利益,因此,有必要考虑多个收费策略建立多目标模型来均衡不同决策者的利益.由于决策者常在信息不确定的情况下做决策,在出行需求不确定的条件下,为了确定次优拥挤收费的方案,建立了基于条件风险价值的随机多目标双层规划模型,上层规划的目标函数考虑了系统总阻抗和社会公平性,下层规划是UE用户均衡配流问题.利用基于随机模拟的遗传算法对模型进行求解,并通过数值算例对模型和算法进行分析,验证了模型的有效性.     

停车收费与拥挤道路收费综合决策研究

停车换乘收费与拥挤道路收费是改善城市交通的有效手段.考虑乘坐公交车,全程自驾车及停车换乘这三种出行方式,给出了停车换乘收费与拥挤道路收费的双层规划模型,证明了模型的一阶条件与出行者的方式选择及路径选择条件等价,并给出了算法.算例分析表明,实行停车收费及拥挤道路收费后,能有效改善道路的拥挤状况,为交通管理部门决策提供了理论依据.     

基于基尼系数的停车换乘拥挤收费定价模型

许多大城市提倡停车换乘的组合出行方式,目的是减少出行车辆,缓解交通拥挤状况.应用双层规划模型对停车换乘拥挤收费进行研究.下层模型采用弹性需求SUE模型,上层模型考虑了交通公平因素.以各出行方式占用的道路资源为公平指标,应用基尼系数对传统上层模型进行改造,构建一个用户盈余尽可能大而基尼系数尽可能小的上层模型,并利用基尼系数控制参数,实现在不同公平要求下的拥挤收费设计.算例表明,基于基尼系数的停车换乘拥挤收费设计,能改善道路的拥挤状况,且兼顾了用户盈余与公平.     

基于出行者属性差异的拥挤收费政策分析

运用经典瓶颈模型,引入出行者属性差异,分析了实行拥挤收费政策前后不同出行方式的出行分布,给出了收费政策影响下出行者行为策略的选择.进一步,分析了基于政府拥挤收费政策下,小汽车和公交车共用同一路线时,道路使用成本、政府调控目标及费率的关系.研究结果表明,政府可以通过设置恰当的收费费率,调节不同出行方式间的出行分布,缓解交通拥挤,优化交通管理现状.属性差异系数的引入使模型更符合实际,改进了已有研究.     

高速公路收费广场的结构设计和管理优化

2017年美国大学生数学建模竞赛B题"Merge After Toll"讨论高速公路收费广场的形状和通行管理方式的优化问题。本文在参考了部分获奖论文基本方法的基础上,对这一问题作了进一步分析,讨论了汇车变道对收费区大小、车流量和安全性的影响,导出了优化收费广场的几个量化指标,并在此基础上分析了收费区设计长度的确定和车流的优化管理方案,给出了相应的仿真计算结果。     

ATIS和道路收费下的混合随机用户均衡的效率损失

在ATIS作用下的交通网络中,用户在交通信息的接受程度上是异质的;考虑到装有ATIS的用户并不总是遵循ATIS的建议,因此,引入信息遵从率这一参数,将所有用户分为三类:安装并遵从信息、安装但不遵从信息和不安装的用户;均按照随机用户均衡的方式进行择路,但对出行时间有不同的感知。同时考虑存在道路收费的情形下,用户在时间价值上是异质的。综合考虑交通信息与道路收费的影响,基于用户两方面的异质性,对其进行合理分类,构建了多用户多准则的混合随机均衡模型及其等价的变分不等式;当收费不作为系统总成本的一部分时,建立了时间准则与费用准则下的系统最优模型;在此基础之上,分别研究了两种准则下混合均衡相对于系统最优的效率损失,给出了效率损失上界,并进一步分析了效率损失上界与各参数间的关系;交通管理者可以从道路收费策略的实施、信息诱导系统的完善等角度进行路网的改造和优化设计。     

多用户多准则随机系统最优与最优收费

针对固定交通需求量和出行者的时间价值为离散分布的多准则随机交通均衡,分别研究了依费用度量和依时间度量的多用户多准则随机系统最优和最优收费问题.分别建立了基于费用和基于时间的随机系统最优的最优化模型,阐述了该模型解的唯一性条件及等价的变分不等式问题.运用变分不等式方法,研究了一阶最优收费的可行性,即能否依边际定价原则,通过收取与出行者类别无关的道路收费使多用户多准则随机均衡流与随机系统最优流一致.一阶最优收费不适用于依时间度量的随机系统最优情况,因而建立了一个最优化模型来得到此时的非歧视性道路收费.最后给出了具体算例.     

加权马尔可夫模型在道路交通受伤人数预测中的应用

为提高交通事故预测的可靠性,首先应用有序聚类的方法建立交通受伤人数的分级标准;然后针对事故受伤人数为相依随机变量的特点,采取以规范化的各阶自相关系数为权重,用加权的马尔可夫链模型来预测未来交通事故受伤人数的升降变化状况;最后以北京市1970-2010年共41年的事故受伤人数为例对该方法进行了具体应用,取得了较为满意的结果,为交通事故受伤人数的预测分析提供了新的途径.     

Highway toll pricing

For a tolled highway where consecutive segments allow vehicles to enter and exit unrestrictedly, we propose a simple toll pricing method. It is shown that the method is the unique method satisfying the classical axioms of Additivity and Dummy in the cost sharing literature, and the axioms of Toll Upper Bound for Local Traffic and Routing-proofness. We also show that the toll pricing method is the only method satisfying Routing-proofness Axiom and Cost Recovery Axiom. The main axiom in the characterizations is Routing-proofness which says that no vehicle can reduce its toll charges by exiting and re-entering intermediately. In the special case when there is only one unit of traffic (vehicle) for each (feasible) pair of entrance and exit, we show that our toll pricing method is the Shapley value of an associated game to the problem. In the case when there is one unit of traffic entering at each entrance but they all exit at the last exit, our toll pricing method coincides with the well-known airport landing fee solution-the Sequential Equal Contribution rule of Littlechild and Owen (1973).     

A Three-Level Mathematical Programming Model of Road Pricing

The paper presents a model of toll imposition on road networks, with the aim of recovering, at least in part, the road costs. This model supposes that the amount of the toll is the result of negotiations between the public agency, which is in charge of the network management, and tends to maximize the toll revenues, and the local communities, who bear the negative consequences of the toll imposition, and are compensated for the damage they sustain. A method of negotiation management is proposed, which compels both parties to behave correctly. The model leads to a three-level hierarchical optimization problem for which a method of solution is proposed and, by way of example, applied to study case.     

Modelling Fuzzy Control Problems with Toll Sets

We shall introduce in this paper toll sets and apply it to dynamical and control systems. We shall reformulate viability theorems for 'fuzzy controlled problem". 'Fuzzification" of dynamical and control systems led us to use indicators instead of characteristic functions for imbedding subsets into functions. Therefore, instead of obtaining the genuine fuzzy sets by a convexification procedure, we obtain the 'toll sets", the membership function being a cost function taking values between 0 and . Hence we can define 'fuzzy" or 'toll" – differential inclusions and control systems and solve the viability problem or the toll viability problem by building fuzzy controllers which regulate viable solutions (satisfying at each instant viability and toll viability constraints).     

Pavement rehabilitation scheduling and toll pricing under different regulatory regimes

This paper addresses the highway pavement rehabilitation scheduling and toll pricing issues over a planning horizon. In the highway system concerned, two types of agents are considered, namely highway operator and road users. Two models, which account for different highway regulatory regimes (i.e. public and private), are proposed. In the public regulatory model, the government aims to maximize total discounted social welfare of the transportation system over the planning horizon by determining the optimal pavement rehabilitation schedule and toll level. In the private regulatory regime, a profit-driven private operator seeks to optimize the pavement rehabilitation schedule and toll level to maximize its own discounted net profit over the planning horizon. The proposed models treat the interactions between the highway operator and the road users in the system as a bi-level hierarchical problem in which the upper level is a multi-period pavement rehabilitation scheduling and toll pricing problem, while the lower level is a multi-period route choice equilibrium problem. A heuristic solution algorithm that combines a greedy approach and a sensitivity analysis based approach is developed to solve the proposed bi-level multi-period optimization models. An illustrative example is used to show the applications of the proposed models. The findings show that the highway regulatory regime, pavement deterioration parameter and the roughness-induced vehicle operating cost can significantly affect the pavement rehabilitation schedules and the toll level as well as the performance of transportation system in terms of total life-cycle travel demand, net profit and social welfare.     

中国物流业碳排放特征及其影响因素分析——基于LMDI分解技术

在对我国物流业碳排放特征进行分析的基础上,运用LMDI分解技术,建立中国物流业人均碳排放的因素分解模型,定量分析了1991-2010年能源结构、能源效率、运输方式、物流发展、经济增长以及人口等6种因素对物流业人均碳排放的影响.分析表明:经济增长是拉动物流业碳排放增长最主要的动力,在研究期间呈指数增长的趋势.运输方式对碳排放增长也表现出明显的促进作用,能源结构和能源效率虽然表现出拉动作用,但效果微弱.而物流发展因素则对物流业人均碳排放量表现出明显的抑制作用.因此,应大力推进物流业科技水平的提高,进一步发挥物流发展因素对碳排放的抑制作用,同时,优化物流运输体系,形成以铁路为主体,水运和管道为支撑,公路和航空运输为辅的低碳化的综合运输体系.     

N-控制策略且温储备失效M/G/1可修排队的可靠性指标

本文研究N-门限值进入控制策略且温储备失效M/G/1可修排队系统, 其中系统在处于温储备失效的状态下最多容许N(1)个顾客进入系统。 运用全概率分解技术和拉普拉斯变换工具, 对服务台第一次失效前的寿命概率分布、不可用度、(0,t]时间内的平均失效次数以及处于温储备失效而等待修理的概率等可靠性指标进行了讨论, 并给出了其稳态结果表达式。 最后, 通过数值实例分析了服务台因温储备故障的稳态不可用度和稳态故障频度随一些参数的变化情况。     

The single vehicle routing problem with toll-by-weight scheme: A branch-and-bound approach

Expressways in China make use of the toll-by-weight scheme, in which expressway tolls are collected based on the weight and traveling distance of the vehicle. Most vehicle routing models assume that the cost of traversing each edge is equivalent to edge length or some constant; as a result, such models cannot be practically applied to the Chinese expressway transportation system. This study addresses a new single vehicle routing problem that takes the vehicle’s (laden and unladen) weight into account. To solve this problem exactly, we provide a branch-and-bound algorithm with a provably valid lower bound measure, along with five dominance checkers for additional pruning. We analyze our algorithm using instances generated from standard TSP test cases, as well as two new sets of test cases based on real expressway information from the Gansu and Jiangxi provinces in China. The algorithm can be applied to any toll scheme in which the toll per unit distance monotonically increases with weight, even if the toll function is non-linear.