Fusion de données pour l'estimation des temps de parcours via la théorie de l'évidence

The emergence of new information technologies and the transformation that has occurred in traffic management have both increased drivers' already considerable need for road traffic information. The travel time is one of the forms in which this information is presented, and a number of systems are based on its dissemination. In this context, this indicator is used as a measure of the impedance (or cost) of routes on the network and/or a congestion indicator. This raises the problem of estimating travel times with an acceptable degree of accuracy, which is a particularly difficult task in urban areas as a result of difficultes of a theoretical, technical and methodological nature. Thus, in order to find out the traffic conditions that prevail on an urban road, the traffic sensors that are usually used to measure traffic conditions are ineffective under certain circumstances. New measurement devices (cameras, GPS or cellphone tracking, etc.) mean that other sources of data are increasingly used in order to supplement the information provided by conventional measurement techniques and improve the accuracy of travel) time estimates. As a result, travel time estimation becomes a typical data fusion problem. This study deals with a multisource estimate of journey times and attempts to provide a comprehensive framework for the utilization of multiple data and demonstrate the feasibility of a travel time estimation system based on the fusion of data of several different types. In this case two types of data are involved, data from conventional induction loop sensors (essentially flow and occupancy measurements) and data from probe vehicles. The selected modelling framework is the Dempster-Shafer Evidence Theory, which has the advantage of being able to take account of both the imprecision and uncertainty of the data. The implementation of this methodology has demonstrated that, in each case, better results are achieved with fusion than with methods based on a single source of data and that the quality of the information, as measured by correctly classified rates, improves as the degree of precision required of the estimate is increased.     

A hybrid model based method for bus travel time estimation

Providing accurate information about bus arrival time to passengers can make the public transport system more attractive. Such information helps the passengers by reducing the uncertainty on waiting time and the associated frustrations. However, accurate estimation of bus travel time is still a challenging problem, especially under heterogeneous and lane-less traffic conditions. The accuracy of such information provided to passengers depends mainly on the estimation method used, which in turns depends on the input data used. Hence, developing suitable estimation methods and identifying the most significant/appropriate input data are important. The present study focused on these aspects of development of estimation methods that can accurately estimate travel time by using significant inputs. In order to identify significant inputs, a data mining technique, namely the k-NN classifying algorithm, was used. It is based on the similarity in pattern between the input and historic data. These identified inputs were then used in a hybrid model that combined exponential smoothing technique with recursive estimation scheme based on the Kalman Filtering (KF) technique. The optimal values of the smoothing parameter were dynamically estimated and were updated using the latest measurements available from the field. The performance of the proposed algorithm showed a clear improvement in estimation accuracy when compared with existing methods.     

Travel time estimation oriented freeway sensor placement problem considering sensor failure

Optimal sensor placement on freeway corridor is of great interest to transportation authorities. However, current traffic sensors are easily subject to various failures. Therefore, it is necessary to incorporate sensor failure into the optimal sensor placement model. In this article, a two-stage stochastic model is proposed for the purpose of travel time estimation on freeway corridor. To balance the effectiveness and reliability, a stochastic conditional value at risk (CVaR) model is also proposed. Since both models are too complicated, a customized genetic algorithm is developed. Numerical experiments show that considering sensor failure makes a significant performance improvement in the sensor placement pattern. Sensitivity analysis is also applied to investigate the impact of a number of allowable sensors and different traffic sensor failure probability.     

Travel mode identification using bluetooth technology

Bluetooth technology has been widely used in transportation studies to collect traffic data. Bluetooth media access control (MAC) readers can be installed along roadways to collect Bluetooth-based data. This data is commonly used to measure traffic performance. One of the advantages of using Bluetooth technology to measure traffic performance is that travel time can be measured directly with a certain level of error instead of by estimation. However, travel time outliers can commonly be observed due to different travel mode on arterials. Since travel mode information cannot be directly obtained from the raw Bluetooth-based data, a mathematical methodology is in need to identify travel mode. In this study, a genetic algorithm and neural network (GANN)-based model was developed to identify travel mode. GPS-enabled devices were used to collect ground truth travel time. In order to additionally compare the model performance, K nearest neighbor (KNN) and support vector machine (SVM) were also implemented. N-fold cross validation was applied to statistically assess the models’ results. Since the model performances depend on the model inputs, seven collections of model inputs were tested in order to achieve the best travel mode identification performance. An arterial segment with four consecutive links and three intersections was selected to be the study segment. The results suggested that correctly identifying the three travel modes successfully every time was not possible, although the GANN based model had low misidentification rates. In our study, 6.12% of autos were misidentified as bikes and 10.53% of bikes were misidentified as autos using three links.     

Developing travel time estimation methods using sparse GPS data

Existing methods of estimating travel time from GPS data are not able to simultaneously take account of the issues related to uncertainties associated with GPS and spatial road network data. Moreover, they typically depend upon high-frequency data sources from specialist data providers, which can be expensive and are not always readily available. The study reported here therefore sought to better estimate travel time using “readily available” vehicle trajectory data from moving sensors such as buses, taxis, and logistical vehicles equipped with GPS in “near” real time. To do this, accurate locations of vehicles on a link were first map-matched to reduce the positioning errors associated with GPS and digital road maps. Two mathematical methods were then developed to estimate link travel times from map-matched GPS fixes, vehicle speeds, and network connectivity information with a special focus on sampling frequencies, vehicle penetration rates, and time window lengths. Global positioning system (GPS) data from Interstate I-880 (California) for a total of 73 vehicles over 6 h were obtained from the University of California Berkeley's Mobile Century Project, and these were used to evaluate several travel time estimation methods, the results of which were then validated against reference travel time data collected from high resolution video cameras. The results indicate that vehicle penetration rates, data sampling frequencies, vehicle coverage on the links, and time window lengths all influence the accuracy of link travel time estimation. The performance was found to be best in the 5-min time window length and for a GPS sampling frequency of 60 s.     

基于元胞自动机的动态路段走行时间计算模型

为建立合理的动态交通网络中路段走行时间模型,分析了动态路段走行时间函数的一般形式,对比国内外常用的几种离散型动态路段走行时间函数,基于元胞自动机交通流模型,建立了动态路段走行时间模型。模型可以根据实际路段驶入率、驶出率,推算出任意时刻进入路段车辆的走行时间,并利用M atlab对模型进行求解和数值分析。结果表明,车辆进入路段后的交通状态是动态路段走行时间的主要影响因素;根据累积驶入驶出车辆数曲线可以直接求出动态路段走行时间,能够为动态交通网络中路径走行时间求解奠定基础。      

Modeling arterial travel time distribution by accounting for link correlations: a copula-based approach

The estimation of urban arterial travel time distribution (TTD) is critical to help implement Intelligent Transportation Systems (ITS) and provide travelers with timely and reliable route guidance. The state-of-practice procedure for arterial TTD estimation commonly assumes that the path travel time follows a certain distribution without considering link correlations. However, this approach appears inappropriate since travel times on successive links are essentially dependent along signalized arterials. In this study, a copula-based approach is proposed to model arterial TTD by accounting for spatial link correlations. First, TTDs on consecutive links along one arterial in Hangzhou, China are investigated. Link TTDs are estimated through the nonparametric kernel smoothing method. Link correlations are analyzed in both unfavorable and favorable coordination cases. Then, Gaussian copula models are introduced to model the dependent structure between link TTDs. The parameters of Gaussian copula are obtained by Maximum-Likelihood Estimation (MLE). Next, path TTDs covering consecutive links are estimated based on the estimated copula models. The results demonstrate the advantage of the proposed copula-based approach, compared with the convolution without capturing link correlations and the empirical distribution fitting methods in both unfavorable and favorable coordination cases.     

Estimation of delay variability at signalized intersections for urban arterial performance evaluation

Urban arterial performance evaluation has been broadly studied, with the major focus on average travel time estimation. However, in view of the stochastic nature of interrupted flow, the ability to capture the characteristics of travel time variability has become a critical step in determining arterial level of service (LOS). This article first presents a stochastic approach that integrates classic cumulative curves and probability theories in order to investigate delay variability at signalized intersections, as a dominant part of the link travel time variability. This serves as a basis for arterial travel time estimation, which can be obtained through a convolution of individual link travel time distributions. The proposed approach is then applied in the estimation of travel time along one arterial in Shanghai, China, with abundant automatic vehicle identification (AVI) data sources. The travel time variability is evaluated thoroughly at 30-min intervals, with promising results achieved in comparison to the field measurements. In addition, the estimated travel time distributions are utilized to illustrate the probability of multiple LOS ranges, namely, reliability LOS. The results provide insights into how we might achieve a more reliable and informative understanding of arterial performance.     

Estimating Missing Traffic Volume Using Low Multilinear Rank Tensor Completion

Traffic volume data have been collected and used for various purposes in some aspects of intelligent transportation systems (ITS) applications. However, the unavoidable detector malfunction can cause data to be missing. It is often necessary to develop an effective approach to recover the missing data. In most previous methods, temporal correlation is explored to reconstruct missing traffic volume. In this article, a new missing traffic volume estimation approach based on tensor completion is proposed by exploring traffic spatial–temporal information. The tensor model is utilized to represent traffic volume, which allows for exploring the multicorrelation of traffic volume in spatial and temporal information simultaneously. In order to estimate the missing traffic volume represented by the tensor model, a novel tensor completion algorithm, called low multilinear rank tensor completion, is proposed to reconstruct the missing entries. The proposed approach is evaluated on the PeMS database. Experimental results demonstrate that the proposed method is more effective than the state-of-art methods, especially when the ratio of missing data is high.     

Neural-network multiple models filter (NMM)-based position estimation system for autonomous vehicles

A highly accurate and reliable vehicle position estimation system is an important component of an autonomous driving system. In generally, a global positioning system (GPS) receiver is employed for the vehicle position estimation of autonomous vehicles. However, a stand-alone GPS does not always provide accurate and reliable information of the vehicle position due to frequent GPS blockages and multipath errors. In order to overcome these problems, a sensor fusion scheme that combines the data from the GPS receiver and several on-board sensors has been studied. In previous researches, a single model filter-based sensor fusion algorithm was used to integrate information from the GPS and on-board sensors. However, an estimate obtained from a single model is difficult to cover the various driving environments, including urban areas, off-road areas, and highways. Thus, a multiple models filter (MMF) has been introduced to address this limitation by adapting multiple models to a wide range of driving conditions. An adaptation of the multiple model is achieved through the use of the model probability. The MMF combines several vehicle models using the model probabilities, which indicate the suitability of the current driving condition. In this paper, we propose a vehicle position estimation algorithm for an autonomous vehicle that is based on a neural network (NN)-based MMF. The model probabilities are determined through the NN. The proposed position estimation system was evaluated through simulations and experiments. The experimental results show that the proposed position estimation algorithm is suitable for application in an autonomous driving system over a wide range of driving conditions.     

Unscented Kalman filter for vehicle state estimation

Vehicle dynamics control (VDC) systems require information about system variables, which cannot be directly measured, e.g. the wheel slip or the vehicle side-slip angle. This paper presents a new concept for the vehicle state estimation under the assumption that the vehicle is equipped with the standard VDC sensors. It is proposed to utilise an unscented Kalman filter for estimation purposes, since it is based on a numerically efficient nonlinear stochastic estimation technique. A planar two-track model is combined with the empiric Magic Formula in order to describe the vehicle and tyre behaviour. Moreover, an advanced vertical tyre load calculation method is developed that additionally considers the vertical tyre stiffness and increases the estimation accuracy. Experimental tests show good accuracy and robustness of the designed vehicle state estimation concept.     

先进的旅行者信息系统对出行者选择行为的影响研究

目前研究先进的旅行者信息系统对出行者选择行为的影响主要集中在对路径选择行为的影响上,而忽略了对出行者出行终点和交通方式选择的影响。假定路网中的出行者一部分装有信息装置,另一部分没有装信息装置,利用离散选择理论中的层次选择结构模型和交通规划理论中的随机均衡方法,研究了先进的旅行者信息系统对出行者终点选择,方式分担和路径选择行为的综合影响,建立了一个与网络均衡条件等价的数学规划模型,设计了模型的求解算法,并用一算例分析了市场渗透率和信息质量对出行者选择行为的影响。     

Development of a behaviorally induced system optimal travel demand management system

The basic design concept of most advanced traveler information systems (ATIS) is to present generic information to travelers, leaving travelers to react to the information in their own way. This “passive” way of managing traffic by providing generic traffic information makes it difficult to predict the outcome and may even incur an adverse effect, such as overreaction (also referred to as the herding effect). Active traffic and demand management (ATDM) is another approach that has received continual attention from both academic research and real-world practice, aiming to effectively influence people's travel demand, provide more travel options, coordinate between travelers, and reduce the need for travel. The research discussed in this article deals with how to provide users with a travel option that aims to minimize the marginal system impact that results from this routing. The goal of this research is to take better advantage of the available real-time traffic information provided by ATIS, to further improve the system level traffic condition from User Equilibrium (UE), or a real-world traffic system that is worse than UE, toward System Optimal (SO), and avoid passively managing traffic. A behaviorally induced, system optimal travel demand management model is presented to achieve this goal through incremental routing. Both analytical derivation and numerical analysis have been conducted on Tucson network in Arizona, as well as on the Capital Area Metropolitan Planning Organization (CAMPO) network in Austin, TX. The outcomes of both studies show that our proposed modeling framework is promising for improving network traffic conditions toward SO, and results in substantial economic savings.     

通行能力约束下的弹性需求交通投影动态演化模型

利用投影动态系统理论建立了具有路段通行能力约束的弹性需求交通网络动态演化模型.通过分析节点路段处交通流量与出行阻抗关系,揭示了出行者在网络局部对出行路线进行调整的决策过程,并分别建立了有通行能力约束的路段流量更新方程和弹性交通需求下的节点最短行程时间估计方程.通过在整个网络上整合上述两类方程,得到最终的交通网络投影动态...     

Updating of Travel Behavior Model Parameters and Estimation of Vehicle Trip Chain Based on Plate Scanning

This article proposes a maximum-likelihood method to update travel behavior model parameters and estimate vehicle trip chain based on plate scanning. The information from plate scanning consists of the vehicle passing time and sequence of scanned vehicles along a series of plate scanning locations (sensor locations installed on road network). The article adopts the hierarchical travel behavior decision model, in which the upper tier is an activity pattern generation model, and the lower tier is a destination and route choice model. The activity pattern is an individual profile of daily performed activities. To obtain reliable estimation results, the sensor location schemes for predicting trip chaining are proposed. The maximum-likelihood estimation problem based on plate scanning is formulated to update model parameters. This problem is solved by the expectation-maximization (EM) algorithm. The model and algorithm are then tested with simulated plate scanning data in a modified Sioux Falls network. The results illustrate the efficiency of the model and its potential for an application to large and complex network cases.     

Real-time queue length estimation using event-based advance detector data

Real-time queue length information at signalized intersections is useful for both performance evaluation and signal optimization. Previous studies have successfully examined the use of high-resolution event-based data to estimate real-time queue lengths. Based on the identification of critical breakpoints, real-time queue lengths can be estimated by applying the commonly used shockwave model. Although breakpoints can be accurately identified using lane-by-lane detection, few studies have investigated queue length estimation using single-channel detection, which is a common detection scheme for actuated signal control. In this study, a breakpoint misidentification checking process and two input-output models (upstream-based and local-based) are proposed to address the overestimation and short queue length estimation problems of breakpoint-based models. These procedures are integrated with a typical breakpoint-based model framework and queue-over-detector identification process. The proposed framework was evaluated using field-collected event-based data along Speedway Boulevard in Tucson, Arizona. Significant improvements in maximum queue length estimates were achieved using the proposed method compared to the breakpoint-based model, with mean absolute errors of 35.7 and 105.6 ft., respectively.     

A framework for travel time variability analysis using urban traffic incident data

This study aims to develop a framework to estimate travel time variability caused by traffic incidents using integrated traffic, road geometry, incident, and weather data. We develop a series of robust regression models based on the data from a stretch in California's highway system during a two-year period. The models estimate highway clearance time and percent changes in speed for both downstream and upstream sections of the incident bottleneck. The results indicate that highway shoulder and lane width factor adversely impact downstream highway clearance time. Next, travel time variability is estimated based on the proposed speed change models. The results of the split-sample validation show the effectiveness of the proposed models in estimating the travel time variability. Application of the model is examined using a micro-simulation, which demonstrates that equipping travelers with the estimated travel time variability in case of an incident can improve the total travel time by almost 60%. The contribution of this research is to bring several datasets together, which can be advantageous to Traffic Incident Management.     

Providing personalized system optimum traveler information in a congested traffic network with mixed users

The advancement of information and communication technology allows the use of more sophisticated information provision strategies for real-time traffic management in a congested network. This article proposes a personalized system optimum traveler information (PSOI) system under ubiquitous communication, which allows traffic system operators to fully optimize and coordinate individuals' trip plans according to the personal attributes, such as real-time location, value of time, allowable budgets for congestion tolling, and willingness to take detours. We also developed an efficient queue-based evaluation and solution heuristic algorithm using mesoscopic simulation models to solve for near-optimal PSOI strategies—route suggestions for each individual traveler. The simulation optimization algorithm can account for different information users and provide predictive information that robustly accounts for potential decisions of other travelers in real time. Case studies were carried out on a test network and a real-world network, and the proposed heuristic algorithm is proven effective. Also, sensitivity analyses show that PSOI not only is an effective traffic management method in reducing average system travel time, but also potentially provides travelers with reasonable or even shorter travel times compared with other information users. Further, simulation results showed that even in mixed traffic, PSOI is able to shorten travel times for both users without information and users of other information types. Thus, PSOI is recommended by this article as an advantageous way for next-generation advanced information systems and dynamic traffic management.     

观测路径出行时间下随机网络交通需求估计

道路网络起讫点（OD）需求是城市决策长期交通规划和短期交通管理中的基础参数,准确的交通需求更是实施交通拥堵控制、限行限速、路径诱导等措施的先决条件。综合运用观测的轨迹已知和未知路径出行时间,建立随机网络交通需求估计双层规划模型。上层广义最小二乘模型最小化历史交通需求与待估交通需求、观测路径出行时间与待估路径出行时间之间的偏差,约束为交通需求、路段流量、路段出行时间与路径出行时间之间的传播关系,通过高斯混合模型（GMM）对其中轨迹未知的观测出行时间依概率聚类。下层为随机网络交通出行均衡模型,分别运用出行时间预算和随机用户均衡处理路网不确定性和出行者感知误差。上、下层之间通过交通需求和OD-路段关联比例进行信息传递。设计迭代算法框架求解双层规划模型,迭代算法包含求解上层模型的最速下降法、求解下层模型的相继平均算法和求解GMM模型的最大期望（EM）算法。通过算例表明轨迹未知的路径出行信息的加入在提升需求估计精度的同时也增大了估计值的方差;设计的迭代算法能够稳定收敛到10^-5的精度;GMM软聚类方法估计的交通需求显著优于硬聚类方法估计的需求值;交通需求值对观测路径出行时间的扰动更加敏感。研究考虑出行者风险态度,通过轨迹信息的重新构建揭示城市交通需求演化规律。