Search

Sensor and machine learning technologies have improved the perception of traffic systems by providing detailed data about individual vehicle trajectories. Combining data from different types of sensors shows promise for comprehensive perception of global traffic, but it remains challenging. Stationary roadside units only gather sparse trajectories of passing vehicles, while crowd-sourced data records entire trajectories but only consists of a very low sample rate of vehicles. Therefore, there is a need to learn route choice behavior from crowd-sourced data to infer complete paths for the sparse trajectories. Existing route choice models assume path set enumeration or the Markovian property for simplicity, which leaves room for capturing the long sequence of choice behavior from data for added precision. Additionally, the path inference problem is often broken down into multiple independent route choice problems between any consecutive sparse observations, leaving room for exploring one-shot long-sequence inference. To address these challenges, we propose RoutesFormer, an efficient sequence-based, data-driven route choice Transformer that requires minimal assumptions due to the capacity of the model architecture. By being sequence-based, RoutesFormer unifies the route choice and path inference problems, accommodating all observations together and avoiding the need to break down the problem into separate route choices, thereby improving optimality. Experiments conducted on the Shanghai taxi dataset demonstrate that RoutesFormer has made significant improvements over six existing baseline models in various challenging path inference tasks. Specifically, RoutesFormer has achieved state-of-the-art accuracy with an average total link length accuracy of 0.914/0.870 compared to the baselines' best average accuracy of 0.896/0.845, and it ranks first across all tasks. Additionally, the attention mechanism used in RoutesFormer is interpreted, providing a lens to study traveler's route choice behavior in the real world. • Innovates a sequence-based path inference model to remedy long-standing drawbacks. • Unifies path inference/route choice in an end-to-end paradigm. • Tailors two types of attention mechanisms to express complex route choice behavior. • Achieves SOTA accuracies, compared with six baselines from classic to deep models. [ABSTRACT FROM AUTHOR]

• Proposes a macro–micro approach for vehicle motion inference using velocity contour. • Macro: tailors an enhanced traffic state estimation method to address data sparsity. • Micro: infers car-following and lane-changing behaviors using a joint optimization framework. • Achieves higher accuracy and smoothness compared to two SOTA methods. Vehicle trajectories can offer the most precise and detailed depiction of traffic flow and serve as a critical component in traffic management and control applications. Various technologies have been applied to reconstruct vehicle trajectories from sparse fixed and mobile detection data. However, existing methods predominantly concentrate on single-lane scenarios and neglect lane-changing (LC) behaviors that occur across multiple lanes, which limit their applicability in practical traffic systems. To address this research gap, we propose a macro–micro approach for reconstructing complete vehicle trajectories on multi-lane freeways, wherein the macro traffic state information and micro driving models are integrated to overcome the restrictions imposed by lane boundary. Particularly, the macroscopic velocity contour maps are established for each lane to regulate the movement of vehicle platoons, meanwhile the velocity difference between adjacent lanes provide valuable criteria for guiding LC behaviors. Simultaneously, the car-following models are extended from micro perspective to supply lane-based candidate trajectories and define the plausible range for LC positions. Later, a two-stage trajectory fusion algorithm is proposed to jointly infer both the car-following and LC behaviors, in which the optimal LC positions is identified and candidate trajectories are adjusted according to their weights. The proposed framework was evaluated using NGSIM and HighD datasets, and the results indicated a remarkable enhancement in both the accuracy and smoothness of reconstructed trajectories, with performance indicators reduced by over 30% compared to two representative reconstruction methods. Furthermore, the reconstruction process effectively reproduced LC behaviors across contiguous lanes, adding to the framework's comprehensiveness and realism. [ABSTRACT FROM AUTHOR]

Traffic volume is an indispensable ingredient to provide fine-grained information for traffic management and control. However, due to the limited deployment of traffic sensors, obtaining full-scale volume information is far from easy. Existing work on this topic focuses primarily on improving the overall estimation accuracy of a particular method and ignores the underlying challenges of volume estimation, thereby having inferior performance in some critical tasks. This paper studies two key problems with regard to traffic volume estimation: (1) underdetermined traffic flows caused by fully undetected paths that can allow arbitrary volume values without violating the conservation law, where using local side information is insufficient to tackle, and (2) nonequilibrium traffic flows arise when traffic flows vary in density over space and time due to congestion propagation delay, which produce time-shifted volume readings to varying degrees. Here we demonstrate a graph-based deep learning method that offers a data-driven, equation-free, and correlation-adaptive approach to address the above issues and perform accurate network-wide traffic volume estimation. Particularly, in order to quantify the dynamic and nonlinear speed-volume relationships for the estimation of underdetermined flows, a speed pattern-adaptive adjacency matrix based on graph attention is developed and integrated into the graph convolution process, to capture nonlocal correlations between volumes. To measure the impacts of nonequilibrium flows, a temporal masked and clipped attention combined with a gated temporal convolution layer is customized to capture time-asynchronous correlations between upstream and downstream sensors under varying impacts of congestion delay. We then evaluate our model on a real-world highway traffic volume dataset and compare it with several benchmark models. It is demonstrated that the proposed model achieves high estimation accuracy even under 20% sensor coverage rate and outperforms other baselines significantly, especially on underdetermined and nonequilibrium flow locations. Furthermore, comprehensive quantitative model analysis is also carried out to justify the model designs. The source code and datasets are publicly available at:. • Identifies underdetermined and non-equilibrium problems for network-wide traffic volume estimation. • Tailors a spatiotemporal correlation adaptive graph convolution network (STCAGCN) for volume estimation. • Exploits global speed-volume relationship as a remedy to underdetermined flows. • Captures time-asynchronous correlations in non-equilibrium flows with masked and clipped attention. • STCAGCN achieves state-of-the-art performances on all tasks. [ABSTRACT FROM AUTHOR]

Highlights • Conversion into a route choice problem enables consideration of traffic variability. • Choice probability model can easily capture effects of non-recurring traffic events. • The model calibrated by massive trip data identifies trip anomalies dynamically. • Unintentional trip anomalies can be separated from the intentional ones effectively. Abstract Taxi service is an essential complement to public transport systems due to its convenience and availability. It often provides hundreds of millions of rides for urban travelers every year in cities across the world. At the same time, the number of trip-induced passenger complaints about trip anomalies (trips with anomalous trip length, time, fare, etc.) is also considerable. Hence, the taxi regulators impose harsh penalties on verified trip anomalies. The existing anomaly verification process is labor-intensive, and it does not consider the traffic variability as well as the passengers' perception of trip anomalies. Quite often the imprecise and unfair outputs are generated as a result. To tackle this issue, we propose a choice probability-based taxi trip anomaly detection model (ProbDetect) that considers the taxi drivers' route choice behavior as well as the traffic variability. We first generate a route choice set for each OD pair based on the massive taxi GPS trajectory data. Second, we assign each route with a choice probability derived from a cumulative multivariate probability over differences of generalized costs. Third, we distinguish the unintentional anomalies from the intentional anomalies using the expected and the realized choice probability. Lastly, the model is tested on 5000 OD pairs using 180 days of taxi GPS data in Shanghai, China. Three types of anomalies are detected as a result. Insights into the driver's route choice behavior are derived as well. [ABSTRACT FROM AUTHOR]

Traffic speed is central to characterizing the fluidity of the road network. Many transportation applications rely on it, such as real-time navigation, dynamic route planning, and congestion management. Rapid advances in sensing and communication techniques make traffic speed detection easier than ever. However, due to sparse deployment of static sensors or low penetration of mobile sensors, speeds detected are incomplete and far from network-wide use. In addition, sensors are prone to error or missing data due to various kinds of reasons, speeds from these sensors can become highly noisy. These drawbacks call for effective techniques to recover credible estimates from the incomplete data. In this work, we first identify the issue as a spatiotemporal kriging problem and propose a Laplacian enhanced low-rank tensor completion (LETC) framework featuring both low-rankness and multi-dimensional correlations for large-scale traffic speed kriging under limited observations. To be specific, three types of speed correlation including temporal continuity, temporal periodicity, and spatial proximity are carefully chosen and simultaneously modeled by three different forms of graph Laplacian, named temporal graph Fourier transform, generalized temporal consistency regularization, and diffusion graph regularization. We then design an efficient solution algorithm via several effective numeric techniques to scale up the proposed model to network-wide kriging. By performing experiments on two public million-level traffic speed datasets, we finally draw the conclusion and find our proposed LETC achieves the state-of-the-art kriging performance even under low observation rates, while at the same time saving more than half computing time compared with baseline methods. Some insights into spatiotemporal traffic data modeling and kriging at the network level are provided as well. • Tailors a Laplacian enhanced low-rank tensor completion (LETC) model for large-scale traffic speed kriging. • Designs a temporal graph Fourier transform considering temporal periodicity for enhancing low-rankness. • Proposes a diffusion graph regularization to exploit spatial dependency on directed graphs. • Develops a generalized temporal consistency regularization to impose smoothness constraints. • LETC achieves state-of-the-art performances with halved computing time. [ABSTRACT FROM AUTHOR]

Intra-driver heterogeneity is defined as transition of driver's behavior between usual and unusual, which is an intrinsic feature of drivers while yet to be extensively explored. This study used a large-scale naturalistic driving data set to investigate intra-driver heterogeneity in car-following. We constructed an IDM-based baseline model to represent a driver's usual behavior; by measuring difference between observed behavior with baseline, unusual behavior was identified. Then, multi-level logit model with random effects was fitted to uncover contributing factors. Among 41 drivers' 1356 trips, intra-driver heterogeneity was identified in 3194 episodes, which accounts for 15% of the time. Within investigated 24 factors, we found that intra-driver heterogeneity was statistically related with vehicle kinematic features, then traffic flow and surrounding environment, but not driver sociodemographics. Being cut in is the most prominent trigger for intra-driver heterogeneity. These findings garner some remarkable insights into improvement of car-following modeling and many other engineering practices. [ABSTRACT FROM AUTHOR]

Taxis provide essential transport services in urban areas. In the taxi industry, the income level remains a cause of concern for taxi drivers as well as regulators. Mining underlying factors affecting the income level will not only benefit the newcomers and low-income drivers but also assist in developing effective optimization algorithms for taxi operations. This paper intends to disclose the factors affecting incomes along with their quantitative influence by mining over 167 million GPS records from nearly 8000 taxis in Shanghai. We first identify a marked difference in drivers’ incomes and categorize drivers into three income levels accordingly. We next investigate the overall search-delivery process, thereby defining several factors that may affect the income level. We then develop a generalized multi-level ordered logit (GMOL) model to find the significant factors that influence incomes. Finally, we compute the elasticity for those significant factors and present their contributions, as well as challenge some preconceived ideas regarding how to earn high incomes. [ABSTRACT FROM AUTHOR]

• Introduces intermodal mobility to tackle imbalance in ride-hailing services. • Maximizes benefits of intermodal ride-hailing services via reinforcement learning. • Tailors a queueing network-based RL policy for added speed of convergence. • Learned policy markedly shortens queue and improves service rates. Growing spatiotemporal imbalance in supply and demand in ride-hailing services has been arousing concern. A variety of measures in the perspective of ride-hailing services per se is studied. However, approaching the imbalance problem in the broader perspective of intermodal mobility, which integrates different modes of passenger transportation in a single trip and aims to overcome limitations of any unimodal mobility, remains to be explored. This paper aims to investigate the potential of introducing intermodal mobility options to balance ride-hailing services. We first identify the importance of the availability decision on intermodal mobility options. Then we formulate the availability decision problem as a Markov decision process (MDP). Due to its convoluted system dynamics and large state space, we cast the intractable MDP into a reinforcement learning (RL) problem to approximately learn the availability policy. To stabilize the learning processes, we model the intermodal ride-hailing services as a stochastic queueing network and tailor a family of state-of-the-art RL algorithms to iteratively evaluate and improve the availability policy. Lastly, we test this optimization framework in a large-scale intermodal mobility scenario calibrated with real-world trip data. Results show that the learned availability policy can significantly dissipate riders' queue and improve the service rates towards more balanced supply and demand. [ABSTRACT FROM AUTHOR]

Rapid advances in sensor, wireless communication, cloud computing and data science have brought unprecedented amount of data to assist transportation engineers and researchers in making better decisions. However, traffic data in reality often has corrupted or incomplete values due to detector and communication malfunctions. Data imputation is thus required to ensure the effectiveness of downstream data-driven applications. To this end, numerous tensor-based methods treating the imputation problem as the low-rank tensor completion (LRTC) have been attempted in previous works. To tackle rank minimization, which is at the core of the LRTC, most of aforementioned methods utilize the tensor nuclear norm (NN) as a convex surrogate for the minimization. However, the over-relaxation issue in NN refrains it from desirable performance in practice. In this paper, we define an innovative nonconvex truncated Schatten p -norm for tensors (TSpN) to approximate tensor rank and impute missing spatiotemporal traffic data under the LRTC framework. We model traffic data into a third-order tensor structure of time intervals × locations (sensors) × days and introduce four complicated missing patterns, including random missing and three fiber-like missing cases according to the tensor mode- n fibers. Despite nonconvexity of the objective function in our model, we derive the global optimal solutions by integrating the alternating direction method of multipliers (ADMM) with generalized soft-thresholding (GST). In addition, we design a truncation rate decay strategy to deal with varying missing rate scenarios. Comprehensive experiments are finally conducted using four different types of real-world spatiotemporal traffic datasets, which demonstrate that the proposed LRTC-TSpN method performs well under various missing cases even with high rates of data loss, meanwhile outperforming other state-of-the-art tensor-based imputation models in almost all scenarios. In general, the performance of the proposed method achieves up to a 52% improvement in mean absolute error (MAE) compared to baseline methods even when the missing ratio is as high as 90%. Moreover, both theoretical and empirical evidences indicate that our method has robust and efficient convergence performances. • A novel truncated tensor Schatten p-norm (TSpN) method is proposed for data imputation. • An efficient solving algorithm is derived with global optimum guarantees. • The TSpN can effectively handle complicated missing patterns even under high missing rates. • The TSpN outperforms other state-of-the-art methods on four real-world traffic data sets. [ABSTRACT FROM AUTHOR]

• Matching time delay and pickup time reduction trade-off unraveled. • Delayed matching outperforms when supply and demand near balance. • Tailored reinforcement learning resolves curse of dimensionality and reward sparsity. • Learned delayed matching policy adaptively reduces total wait costs. Matching trip requests and available drivers efficiently is considered a central operational problem for ride-hailing services. A widely adopted matching strategy is to accumulate a batch of potential passenger-driver matches and solve bipartite matching problems repeatedly. The efficiency of matching can be improved substantially if the matching is delayed by adaptively adjusting the matching time interval. The optimal delayed matching is subject to the trade-off between the delay penalty and the reduced wait cost and is dependent on the system's supply and demand states. Searching for the optimal delayed matching policy is challenging, as the current policy is compounded with past actions. To this end, we tailor a family of reinforcement learning-based methods to overcome the curse of dimensionality and sparse reward issues. In addition, this work provides a solution to spatial partitioning balance between the state representation error and the optimality gap of asynchronous matching. Lastly, we examine the proposed methods with real-world taxi trajectory data and garner managerial insights into the general delayed matching policies. The focus of this work is single-ride service due to limited access to shared ride data, while the general framework can be extended to the setting with a ride-pooling component. [ABSTRACT FROM AUTHOR]

Adverse weathers are well-known to impact the operation of transportation systems, including taxis. This paper utilizes taxi GPS waypoint data to investigate the quantitative impact of rainfall on taxi hailing and taxi operations to help improve service quality on rainy days. Through statistical analysis, the study proves that it is more difficult to hail taxis on rainy days, especially during morning peak hours. By modelling the difference value of factors for rainfall and nonrainfall conditions in a multivariate regression model and attaining the significance and elasticity of each factor, passenger demand, taxi supply, search time and velocity are proved to be the significant factors that lower the taxis' level of service on rainy days. Among them, the number of passengers and taxis are two factors that have the greatest impact. It is also shown that there is no significant difference in the total taxi supply and passenger demand between rainfall and nonrainfall conditions, but a dramatic change in the spatial distribution is discovered. The results suggest that instead of simply providing more taxis on rainy days, optimally dispatching taxicabs to high demand regions can be a more effective solution. [ABSTRACT FROM AUTHOR]

• Integrates macro–micro models to reconstruct full-sampled trajectories using sparse data. • Macro: applies shockwave and fundamental diagram to infer velocity bounds. • Micro: extends car-following models to generate valid trajectory candidates. • Solves optimal fusion of trajectory candidates as per inferred velocity bounds. • Test on NGSIM datasets shows notable improvements in accuracies and smoothness. Vehicle trajectories can provide a clear picture of the traffic flow that plays a pivotal role in traffic management and control. Two types of traffic sensors, i.e., fixed and mobile sensors, are widely adopted to collect road traffic information. However, fully-sampled vehicle trajectories are far from obtainable through those sensors due to their fixed-location measurements or low penetration rates. To overcome these issues, this study proposes an integrated macro–micro framework to reconstruct individual vehicle trajectories on freeways under a multi-source data environment, in which fixed-location sensor data that collect positions and velocities of vehicles in a certain place and a small fraction of probe vehicle (PV) data that yield continuous trajectories are available. The framework consists of three modules. First, at the macro level, we apply shockwave theory and fundamental diagram to expand the partially detected information to the whole space–time diagram, and then establish a velocity baseline for evaluation of the trajectory fusion thereafter. Second, at the micro level, we develop two trajectory estimation algorithms based on extended car-following models to generate individual trajectories between any two consecutive PVs, in which two candidate trajectories will be generated for each non-probe vehicle. Third, we tailor a trajectory fusion algorithm to integrate the macro- and micro-level models, and apply dynamic programming to solve the optimal trajectories as per the established velocity baseline. The proposed framework is finally tested with NGSIM and HighD datasets. The results show that our integrated method attains a notable improvement in accuracies and smoothness of reconstructed trajectories compared to the non-integrated and the classical Variational Theory-based methods respectively. [ABSTRACT FROM AUTHOR]