The CONTRAM Dynamic Traffic Assignment Model

CONTRAM is a computer model of time-varying traffic in road networks, which takes as input the network definition and time-varying demand for travel between a set of origin and destination zones, and outputs the resulting network flows, routes and travel times. It combines a macroscopic time-sliced traffic model with disaggregate dynamic assignment of traffic, so is intermediate between macroscopic equilibrium and microscopic models. The paper details the methods used, including time-dependent queuing which plays a central role, and the treatment of network definition, user classes, road capacities, signals and coordination, vehicle emissions, Intelligent Transport Systems and research lines.     

出租车数据的城市道路网路段通行时间估计方法

城市路段通行时间估计能够更好地运营和管理城市交通。针对包含起点-终点位置,行程时间和距离信息的GPS行程数据,提出了一种城市道路网短时通行时间的估计模型。首先将城市道路网按照交叉路口分解为多个路段,并基于k-最短路径搜索方法分析司机行进路线。然后针对每一个路段,提出了双车道通行时间多项式关联关系模型,既能提升道路网通行时间精细度,又能避免因训练数据不足导致的路网通行时间过拟合问题。最后以最小化行程期望时间和实际行程时间之间的均方误差为优化目标,拟合道路网通行时间。在纽约出租车数据集上的实验结果表明,所提模型及方法相对于传统单车道估计方法能够更准确地估计城市道路网路段的通行时间。     

Estimating congestion zones and travel time indexes based on the floating car data

Efficiently predicting traffic congestion benefits various traffic stakeholders, from regular commuters and logistic operators to urban planners and responsible authorities. This study aims to give a high-quality estimation of traffic conditions from a large historical Floating Car Data (FCD) with two main goals: (i) estimation of congestion zones on a large road network, and (ii) estimation of travel times within congestion zones in the form of the time-varying Travel Time Indexes (TTIs). On the micro level, the traffic conditions, in the form of speed profiles were mapped to links in the road network. On the macro level, the observed area was divided into a fine-grained grid and represented as an image where each pixel indicated congestion intensity. Spatio-temporal characteristics of congestion zones were determined by morphological closing operation and Monte Carlo simulation coupled with temporal clustering. As a case study, the road network in Croatia was selected with spatio-temporal analysis differentiating between the summer season and the rest of the year season. To validate the proposed approach, three comparisons were conducted: (i) comparison to real routes' travel times driven in a controlled manner, (ii) comparison to historical trajectory dataset, and (iii) comparison to the state-of-the-art method. Compared to the real measured travel times, using zone's time-varying TTIs for travel time estimation resulted in the mean relative percentage error of 4.13%, with a minor difference to travel times estimated on the micro level, and a significant improvement compared to the current Croatian industrial navigation. The results support the feasibility of estimating congestion zones and time-varying TTIs on a large road network from FCD, with the application in urban planning and time-dependent routing operations due to: significant reduction in the data volume without notable quality loss, and meaningful reduction in the pre-processing computation time.     

时间依赖路网中反向k近邻查询

在现存的反向k近邻查询方案中,比较高效的研究大多集中在欧氏空间或者静态路网,对时间依赖路网中的反向k近邻查询的研究相对较少。已有算法在兴趣点密度稀疏或者k值较大时,查询效率较低。对此,提出了基于子网划分的反向k近邻查询算法mTD-SubG。首先,将整个路网划分为大小相同的子网,通过子网的边界节点向其他子网进行扩展,加快对路网中兴趣点的查找速度;其次,利用剪枝技术缩小路网的扩展范围;最后, 利用已有时间依赖路网下的近邻查询算法,判定查找到的兴趣点是否为反向k近邻结果。实验中将mTD-SubG算法与已有算法mTD-Eager进行对比,结果表明mTD-SubG算法的响应时间比mTD-Eager算法减少了85.05%,遍历节点个数比mTD-Eager算法减少了51.40%。     

Reliability Evaluation for Stochastic and Time-dependent Networks with Multiple Parking Facilities

This paper presents a fixed-point model for evaluating the reliability of stochastic and time-dependent networks with multiple parking facilities. The proposed model combines a supply model that simulates the time-dependent attributes of road and parking supplies and their variations with a demand model that simultaneously considers heterogeneous travelers’ choices on departure time, route and parking location. In the proposed model, travelers are differentiated by their values of time, and parking locations are characterized by facility type. Schedule reliability and parking reliability are introduced as new performance indices for the evaluation of the level of service of a road network during time of day. A heuristic solution algorithm that uses a combination of the Monte Carlo simulation approach with the method of successive averages is proposed to estimate these two reliability measures. Numerical results show that the reliability performances of road networks are significantly influenced by the network congestion level and the capacities of road and parking facilities. The proposed model provides some new insights for assessing the impacts of various transport policies and infrastructure improvements at a strategic level.     

用于热门路径查询的动态热度路网的构建方法

基于位置的服务在蓬勃发展的同时,产生出大量的用户位置轨迹数据,同时基于轨迹数据的热门路径问题也越来越受到人们的重视。对于求解点到点热门路径的问题,一张带有动态热度信息的热度路网是非常必要的。首先,提出了一个高效的交叉口生成算法,用于构建静态的热度路网,并在此基础上提出一种新的时间段分割算法来使得路网中对应热度边动态化。采用希腊雅典的一部分卡车的GPS轨迹数据集,通过大量充分的实验,印证了算法的合理性和高效性。     

基于Katz自动编码器的城市路网链路预测模型

城市交通道路网络（以下简称“路网”）是一种特殊的复杂网络,对路网进行链路预测在城市规划与城市结构演化方面有着重要的应用价值。针对路网的高度稀疏性、高度非线性特点,提出了一种基于Katz相似度自动编码器（Katz Auto Encoder Network Embedding,KAENE）的路网链路预测模型,它是一种基于自动编码器的深度学习网络嵌入模型,使用Katz相似度矩阵保存路网的结构特征,利用多层非线性自动编码器对路网进行网络表征学习,在模型训练阶段通过局部线性嵌入损失函数保存路网的局部特征,在此基础上引入L2范数来提高模型的泛化能力,最后结合路网的方向性特征提高路网的链路预测精确度。通过实验对比了KAENE模型与其他链路预测模型在国内外的不同城市路网数据上的表现以及不同嵌入维度对KAENE模型预测精度的影响,最后通过可视化了解了模型的网络表征学习过程。实验结果表明,KAENE在国内外6个具有代表性的路网数据集的链路预测任务中取得了良好的表现。     

基于改进遗传算法的动态路径规划研究

在静态路网模型的基础上构建时间依赖的动态路网模型数据库,进行动态路径规划问题研究。针对传统遗传算法在解决此问题中存在的“早熟收敛”、局部搜索能力差等问题,对其进行下列改进：结合随机选择和趋于终点方向的种群初始化策略,在保持初始种群多样性的同时提高其个体质量;根据空间邻近关系选择交叉位置点,有效保留父代优良基因,同时避免“早熟收敛”;采用节点适应度的局部搜索策略,根据路段所属道路等级、转弯类型、实时路况以及与局部路段终点的夹角四个影响因子,构建当前节点邻接节点的适应度,提高局部搜索能力。研究结果表明,改进后的遗传算法具有更好的收敛效果和收敛稳定性,满足行进中的动态最优路径规划对求解精度和效率的要求。     

The time-dependent prize-collecting arc routing problem

A new problem is introduced named the Time-Dependent Prize-Collecting Arc Routing Problem (TD-PARP). It is particularly relevant to situations where a transport manager has to choose between a number of full truck load pick-ups and deliveries on a road network where travel times change with the time of day. Two metaheuristic algorithms, one based on Variable Neighborhood Search and one based on Tabu Search, are proposed and tested for a set of benchmark problems, generated from real road networks and travel time information. Both algorithms are capable of finding good solutions, though the VNS approach generally shows better performance.     

最小时间路径算法的改进及在路径优化中的应用*

由于城市交通网络中路径行程时间是随着时间的变化而变化的,求解最小时间路径比较困难,为此提出把交通网络抽象为时间依赖的网络模型的解决方法。对时间依赖网络模型和理论基础进行分析,指出文献［1］描述的最小时间路径算法存在的不足,即不能正确记录路径;通过引入一个记录路径的数组来对此算法进行改进,改进后的算法不仅解决了原算法存在的问题,而且可以满足n∶1的最短路径搜索,扩展了原算法的应用范围。最后用实验验证了改进算法的正确性和有效性。     

基于最优方案VANET协议的研究

由于网络拓扑结构频繁变化,节点之间物理物理距离超过通信距离时,消息不能转发,信息没有及时更新。目前市场销售的大部分导航软件一般采用的是最短路径策略或最少收费策略,这种方式在一般情况下能够满足人们的某一出行要求,但对于当前交通路况拥堵的情况下,这几种策略显然是行不通的,最短路径或最少收费的方式不仅不能给使用者节省时间,还可能降低了出行的效率。这种方式以远远不能满足人们出行的需求,如何保障即尽可能选择最短路径又能够保证通信的畅通,是本论文研究的方向。     

时间依赖路网上的移动对象K近邻查询算法

随着基于位置服务的广泛应用,时间依赖路网上的对象查询逐渐成为研究热点。以往研究大多只针对时间依赖路网上的静态对象(如加油站、餐厅等),未考虑到移动对象(如出租车)的情况,而移动对象的查询在日常生活中有着非常广泛的应用场景。因此,文中提出了一种针对时间依赖路网上的移动对象K近邻查询算法TD-MOKNN,该算法分为预处理阶段和查询阶段。在预处理阶段,通过建立路网和网格索引,提出了一种新的移动对象到路网的映射方法,解除了以往研究假设移动对象恰好在路网顶点上的限制;在查询阶段,采用启发式搜索,借助倒排网格索引计算了一种新的高效启发值,通过预处理信息和启发值设计了高效K近邻查询算法,并给出了算法的正确性证明和时间复杂度分析。实验验证了所提算法的有效性,相比现有算法,TD-MOKNN算法在遍历顶点数和响应时间上分别减少了55.91%和54.57%,查询效率平均提升了55.2%。     

道路转向延迟的动态对偶图模型

传统的道路转向延迟对偶图表达法缺乏对交通网络时间依赖特性的考虑,不适合动态路径规划问题的求解。本文将时间因素引入到对偶图中,发展了一种动态对偶图模型,将交通路网表达为动态对偶网络,并为之定义了FIFO(先进先出）条件,推导了满足FIFO条件的动态行程计算方法,设计了时间依赖的标号设定最短路径算法。实验结果表明,利用该对偶图模型和动态对偶网络,能有效表达路网转向延迟,在以出行时间为标准的动态路径规划中,基于动态对偶网络的路径规划结果可节省约16%的出行时间。     

基于公共交通的高精度时空可达模型与算法

公共交通作为我国城市居民的主要出行方式,对其可达性的研究具有非常重要的价值和意义.然而,由于站点位置、固定线路、时刻表等条件的限制,使得公共交通可达性的研究具有一定的特殊性.针对已有研究存在的可达精度不高或者无法进行大规模可达分析等问题,基于路网、公交网络、地铁网络和时刻表信息建立高精度时空网络模型,设计时间依赖条件下枢纽站点和A^*算法相结合的快速公交换乘算法.以武汉市为例,对其进行大规模高精度时空可达分析,证明了模型的可靠性和算法的高效性.     

高分辨率多光谱遥感影像中城区道路信息的自动提取

提出一种从高分辨率遥感影像提取城市区域道路网络的方法。该方法采用改进的数学形态学运算方法对影像进行分割,进而得到粗略道路信息网,然后利用道路网的几何特征实现道路与建筑物的有效区分,最后通过抽骨架的方法获得最终道路网中心线。试验数据为某一城区高分辨率卫星影像,并对最终提取的结果进行了评价,结果表明,所提出的方法能够从高分辨率多波段卫星遥感影像上精确、有效、自动提取城区道路网络。     

SMashQ: spatial mashup framework for k-NN queries in time-dependent road networks

The k-nearest-neighbor (k-NN) query is one of the most popular spatial query types for location-based services (LBS). In this paper, we focus on k-NN queries in time-dependent road networks, where the travel time between two locations may vary significantly at different time of the day. In practice, it is costly for a LBS provider to collect real-time traffic data from vehicles or roadside sensors to compute the best route from a user to a spatial object of interest in terms of the travel time. Thus, we design SMashQ, a server-side spatial mashup framework that enables a database server to efficiently evaluate k-NN queries using the route information and travel time accessed from an external Web mapping service, e.g., Microsoft Bing Maps. Due to the expensive cost and limitations of retrieving such external information, we propose three shared execution optimizations for SMashQ, namely, object grouping, direction sharing, and user grouping, to reduce the number of external Web mapping requests and provide highly accurate query answers. We evaluate SMashQ using Microsoft Bing Maps, a real road network, real data sets, and a synthetic data set. Experimental results show that SMashQ is efficient and capable of producing highly accurate query answers.     

An incremental neural learning framework and its application to vehicle diagnostics

This paper presents a framework for incremental neural learning (INL) that allows a base neural learning system to incrementally learn new knowledge from only new data without forgetting the existing knowledge. Upon subsequent encounters of new data examples, INL utilizes prior knowledge to direct its incremental learning. A number of critical issues are addressed including when to make the system learn new knowledge, how to learn new knowledge without forgetting existing knowledge, how to perform inference using both the existing and the newly learnt knowledge, and how to detect and deal with aged learnt systems. To validate the proposed INL framework, we use backpropagation (BP) as a base learner and a multi-layer neural network as a base intelligent system. INL has several advantages over existing incremental algorithms: it can be applied to a broad range of neural network systems beyond the BP trained neural networks; it retains the existing neural network structures and weights even during incremental learning; the neural network committees generated by INL do not interact with one another and each sees the same inputs and error signals at the same time; this limited communication makes the INL architecture attractive for parallel implementation. We have applied INL to two vehicle fault diagnostics problems: end-of-line test in auto assembly plants and onboard vehicle misfire detection. These experimental results demonstrate that the INL framework has the capability to successfully perform incremental learning from unbalanced and noisy data. In order to show the general capabilities of INL, we also applied INL to three general machine learning benchmark data sets. The INL systems showed good generalization capabilities in comparison with other well known machine learning algorithms.     

Data-driven approaches for emissions-minimized paths in urban areas

Concerns about air quality and global warming have led to numerous initiatives to reduce emissions. In general, emissions are proportional to the amount of fuel consumed, and the amount of fuel consumed is a function of speed, distance, acceleration, and weight of the vehicle. In urban areas, vehicles must often travel at the speed of traffic, and congestion can impact this speed particularly at certain times of day. Further, for any given time of day, the observations of speeds on an arc can exhibit significant variability. Because of the nonlinearity of emissions curves, optimizing emissions in an urban area requires explicit consideration of the variability in the speed of traffic on arcs in the network. We introduce a shortest path algorithm that incorporates sampling to both account for variability in travel speeds and to estimate arrival time distributions at nodes on a path. We also suggest a method for transforming speed data into time-dependent emissions values thus converting the problem into a time-dependent, but deterministic shortest path problem. Our results demonstrate the effectiveness of the proposed approaches in reducing emissions relative to the use of minimum distance and time-dependent paths. In this paper, we also identify some of the challenges associated with using large data sets.     

Variances of link travel time estimates: implications for optimal routes

In this paper, we explore the consequences of using link travel time estimates with high variance to compute the minimum travel time route between an origin and destination pair. Because of platoon formation or for other reasons, vehicles on a link separated by small headways tend to have similar travel times. In other words, the covariance of link travel times of distinct vehicles which are close together may not be zero. It follows that the variance of the mean of travel times obtained from a sample of n vehicles on a same link over small time intervals is of the form a + b / n where a and b would usually be positive. This result has an important implication for the quality of road network travel time information given by Intelligent Transportation Systems (ITS)—that the variance of the estimate of mean travel time does not go to zero with increasing n . Thus the quality of information disseminated by ITS is not necessarily improved by increasing the market penetration of vehicles monitoring the system with the necessary equipment (termed probe vehicles). Estimates of a and b for a set of links are presented in the paper and consequences for probe-based ITS are explored by means of a simulation of such a system which is operational on an actual network.     

面向车载容迟网络的连通性建模与仿真研究

针对车载容迟网络连通性建模进行了研究。首先假设车辆驶入道路的过程服从泊松分布,以及车辆在道路上的行驶速度服从正态分布。继而对基于泊松过程的车间时距分布进行推导,并以此导出行驶车辆在道路上的连通概率。为了验证所提假设和连通模型的正确性和有效性,以欧洲城市卢森堡在7:30 a.m.~8:30 a.m.时间段内的交通数据为实验场景,在城市交通仿真平台（simulation of urban mobility,SUMO）对车辆速度的概率分布、车辆到达率、道路中的平均车辆数及网络连通概率进行了理论计算和仿真实验分析。实验结果表明理论模型的计算值和仿真结果是一致的,所提出的假设和连通模型具有合理性和正确性。