Publications
75 results found
Wang Y, Yu X, Guo J, et al., 2022, Macroscopic traffic flow modelling of large-scale freeway networks with field data verification: State-of-the-art review, benchmarking framework, and case studies using METANET, Transportation Research Part C: Emerging Technologies, Vol: 145, ISSN: 0968-090X
Macroscopic traffic flow models are of paramount importance to traffic surveillance and control. Before their employments in applications, the models need to be calibrated and validated against real traffic data. The model calibration determines an optimal set of model parameters that minimizes the discrepancy between the modeling results and real traffic data. The model validation is furthermore performed to corroborate the accuracy of a calibrated model using data other than used for calibration. The model calibration aims to reflect traffic reality, while model validation focuses on the prediction of future traffic using calibrated models. This paper delivers a comprehensive review of state-of-the-art works on macroscopic model calibration and validation, proposes a benchmarking framework on traffic flow modeling, and has conducted a large number of case studies based on the framework using macroscopic traffic flow model METANET with respect to the urban expressway network in Shanghai. In comparison to previous works, quite more comprehensive results on model calibration have been presented in this paper, in consideration of congestion tracking, traffic flow inhomogeneity, capacity drop, stop-and-go waves, scattering, adverse weather conditions, and accidents. The paper has also reported many results of model validation with respect to the same field examples. The results demonstrate that METANET is able to model complex traffic flow dynamics in large-scale freeway networks with sufficient accuracy. The paper is closed with discussion on limitations and future works.
Li Y, Chen B, Zhao H, et al., 2022, A Car-Following Model for Connected and Automated Vehicles With Heterogeneous Time Delays Under Fixed and Switching Communication Topologies, IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS, Vol: 23, Pages: 14846-14858, ISSN: 1524-9050
- Author Web Link
- Cite
- Citations: 1
Yang H, Bao Y, Huo J, et al., 2022, Impact of road features on shared e-scooter trip volume: A study based on multiple membership multilevel model, Travel Behaviour and Society, Vol: 28, Pages: 204-213, ISSN: 2214-367X
E-scooter sharing systems have been widely adopted by cities around the world. Previous studies analyzed community-level factors influencing e-scooter usage. Few studies examined the effect of road features on e-scooter trip volume (ETV) of the road segment, which can reveal the road features that riders prefer. This study explores this topic by analyzing the ETV of 29,544 road segments in Calgary, Canada, while controlling for community-level factors. Because some segments are the boundaries of multiple communities, the multiple membership multilevel model is adopted to tackle this boundary problem. The results show that segments with sidewalks, dedicated bicycle facilities, lower speed limit, more street lights and trees have higher ETV. ETV is also higher in communities with high income, high percentage of commercial and residential area. Quantifying the effect of road features on ETV could help government agencies determine where e-scooters should be ridden and design road facility improvement plans for e-scooter users.
Li J, Xie N, Zhang K, et al., 2022, Network-scale traffic prediction via knowledge transfer and regional MFD analysis, TRANSPORTATION RESEARCH PART C-EMERGING TECHNOLOGIES, Vol: 141, ISSN: 0968-090X
- Author Web Link
- Cite
- Citations: 2
Hu S, Zhou Q, Li J, et al., 2022, High Time-Resolution Queue Profile Estimation at Signalized Intersections Based on Extended Kalman Filtering, IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS, Vol: 23, Pages: 21274-21290, ISSN: 1524-9050
Li Y, Lv Q, Zhu H, et al., 2022, Variable Time Headway Policy Based Platoon Control for Heterogeneous Connected Vehicles With External Disturbances, IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS, Vol: 23, Pages: 21190-21200, ISSN: 1524-9050
Li Y, Zhong Z, Song Y, et al., 2022, Longitudinal Platoon Control of Connected Vehicles: Analysis and Verification, IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS, Vol: 23, Pages: 4225-4235, ISSN: 1524-9050
- Author Web Link
- Cite
- Citations: 2
Shu S, Chen Z, Yu Z, et al., 2022, Modeling Freight-Sharing Platform Operations for Optimal Compensation Strategy Using Markov Decision Processes, Pages: 1006-1011
The urban freight-sharing is one special part of ride-sharing due to its characteristics corresponding to urban freight orders, including the high fragmentation of both demand and supply and geographic concentration. This study has applied a Markov Decision Process framework to determine the static and dynamic optimal compensation strategy offered to shippers and carriers, which aims to maximize the longterm accumulated expected discounted rewards for the freight-sharing platform. More specifically, with the incorporation of stochastic arrival of shippers and carriers, decisions of a shipper placing an order and a carrier accepting an order, the maximum amount of orders and carriers the platform could accommodate, and the current state of the platform regarding the number of unmatched orders and carriers, models are designed to give insights about the optimal compensation-settings under various scenarios with different supply and demand arrival rate. The developed models are tested with the real-world data.
Zhang K, Li J, Zhou Q, et al., 2022, Short-Term Traffic Prediction with Balanced Domain Adaptation, Pages: 699-711
Short-term traffic forecasting has been a hot topic in the intelligent transportation systems field. The traditional traffic forecasting methods mostly fix traffic sensors. However, most sensors are subject to bad conditions, leading to noisy and insufficient raw data. Recent advances have provided new traffic prediction opportunities. For example, the transfer learning method takes advantage of data trained on one good dataset and transfers the knowledge to others with bad data. Existing applications do not consider the underlying data distributions sufficiently, limiting the prediction performance. We propose a transfer learning-based traffic flow prediction framework using the Balanced Domain Adaptation (BDA) method. Various regression models are fed into the framework to evaluate a good data source and predict bad target datasets. A case study using data from the Highways England is conducted. The results show that the proposed BDA-based framework can match the distributions between traffic flow datasets and significantly improve prediction accuracy.
Zhu Y, Li Y, Hu S, et al., 2022, Optimal Control for Vehicle Platoon Considering External Disturbances, Pages: 453-458
This article designs a novel controller that is composed of a linear quadratic regulator (LQR)-based controller and a sliding mode controller (SMC) to guarantee the consensus of vehicle platoon considering the effect of external disturbances. Particularly, a LQR-based controller is designed to ensure the consensus related to position, velocity and acceleration of vehicles in the platoon and the optimization of performance under a diversity of communication topologies. In addition, a SMC is used to deal with external disturbances. Under the LQR-based controller, the stability of the designed controller and the effects of communication topology are analyzed by the Lyapunov method. Finally, the simulation experiments illustrate the superiority of our proposal.
Zhou C, Xiao D, Hu J, et al., 2022, An Example of Digital Twins for Bridge Monitoring and Maintenance: Preliminary Results, Pages: 1134-1143, ISBN: 9783030918767
In this study, we propose a digital twin pilot study for bridge monitoring and maintenance. In particular, an infrastructure management framework using UAV and surveillance cameras, and accelerometers-based digital twins is proposed to perform long-term and non-interruptive monitoring. Real-world monitoring data are obtained through an experimental test performed on the Juanhu bridge (Haning, Zhejiang, China). Traffic flow and accelerometer data of the tested bridge were measured. The digital twin model of the bridge is created as a real-time Finite Element model in OpenSees. The FE model geometry is produced using a 3D photogrammetric reconstruction, and its dynamic properties are updated based on Bayesian modal identification. The traffic flow information on the bridge is processed through computer vision techniques using the video footage from the UAV and surveillance cameras. The object detection algorithm YOLO and tracking algorithm DeepSORT are used to derive the time-space diagrams. These elements operate in tandem with the accelerometer data and the digital twin FE model to acquire a preliminary vehicle loading estimation. The results are presented in this study and showcase the feasibility of the proposed digital twin framework for bridge monitoring and maintenance.
Ye A, Zhou Q, Liu X, et al., 2022, Modeling and Managing an On-Demand Meal Delivery System with Mixed Autonomy, Pages: 2007-2012
This paper investigates the on-demand meal delivery system with mixed autonomy. We have explored how the future implementation of autonomous vehicles (AVs) in the system will affects demand, the labor market of human couriers (HCs), and the service provider in the system. In the system, the service provider determines the fleet size of AVs, the average delivery price for customers, and the average hourly wage for HCs. In response to the operation and pricing strategies, customers decide whether or not to order meals with delivery services, and potential HCs decide whether or not to work for the system. Therefore, a market model is proposed to capture the interactions among the service provider, customers, and HCs. An adaptive particle swarm optimization (APSO) algorithm is adopted to find optimal solutions. The results of numerical experiments show that a lower cost of AVs leads to higher penetration of AVs, lowered delivery price, and improved service quality. As a result, expanded demand is expected. By comparing the market outcomes under a varying number of potential customers, we find that AVs are considered more cost-efficient in densely populated areas than HCs, and have a higher percentage in the mixed fleet. Henee, customers in those areas are served with improved quality of delivery services.
Wang Y, Zhao M, Yu X, et al., 2022, Real-time joint traffic state and model parameter estimation on freeways with fixed sensors and connected vehicles: State-of-the-art overview, methods, and case studies, TRANSPORTATION RESEARCH PART C-EMERGING TECHNOLOGIES, Vol: 134, ISSN: 0968-090X
- Author Web Link
- Cite
- Citations: 3
Liu W, Sun H, Lai D, et al., 2022, Performance of fast fluid dynamics with a semi-Lagrangian scheme and an implicit upwind scheme in simulating indoor/outdoor airflow, BUILDING AND ENVIRONMENT, Vol: 207, ISSN: 0360-1323
- Author Web Link
- Cite
- Citations: 2
Hu S, Shu S, Bishop J, et al., 2022, Vehicle telematics data for urban freight environmental impact analysis, Transportation Research Part D: Transport and Environment, Vol: 102, Pages: 103121-103121, ISSN: 1361-9209
Road freight transport is one of the major contributors to greenhouse gas and air pollutantemissions. Hence, it is increasingly regulated in urban areas to reduce its impact on theenvironment and human health. The rich data available from telematics has the potentialto provide high-resolution information, yet research has not been conducted to understand,evaluate, and ultimately improve the operation and impacts of urban road freight. This paperdemonstrates the role of vehicle telematics data in enabling quantitative assessment of theimpacts of urban freight transport for the effective management of relevant policies. We presenta comprehensive data-driven approach that provides a robust quantitative evaluation andapplies it to a case study of the London Lorry Control Scheme (LLCS) policy in UK. We showthat, for the studied freight operator, the LLCS policy affects their drivers’ route choice bothinside and outside the restricted hours. The spatio-temporal distributions of different parametersincluding traffic speeds, fuel economy and emissions at different times of the day are comparedand analyzed. The results indicate that the unintended consequences of urban freight transportpolicies can include an extra 15% vehicle-km traveled per trip and 12% liter of fuel consumedper trip
Chen C, Hu S, Ochieng WY, et al., 2021, Understanding City-Wide Ride-Sourcing Travel Flow: A Geographically Weighted Regression Approach, JOURNAL OF ADVANCED TRANSPORTATION, Vol: 2021, ISSN: 0197-6729
Liu Q, Hu S, Angeloudis P, et al., 2021, Dynamic wireless power transfer system for electric-powered connected and autonomous vehicle on urban road network, IET INTELLIGENT TRANSPORT SYSTEMS, Vol: 15, Pages: 1153-1166, ISSN: 1751-956X
- Author Web Link
- Cite
- Citations: 3
Tang C, Hu W, Hu S, et al., 2021, Urban traffic route guidance method with high adaptive learning ability under diverse traffic scenarios, IEEE Transactions on Intelligent Transportation Systems, Vol: 22, Pages: 2956-2968, ISSN: 1524-9050
With the rapid development of urbanization, the problem of urban traffic congestion has become increasingly prominent. Dynamic route guidance promises to improve the capacity of urban traffic management and mitigate traffic congestion in big cities. In the design of simulation-based experiments for most dynamic route guidance methods, the simulation data is generally estimated from a specific traffic scenario in the real-world. However, highly dynamic traffic in the city implies that traffic scenarios in real systems are diverse. Therefore, if a route guidance method cannot adjust its strategy according to the spatial and temporal characteristics of different traffic scenarios, then it cannot guarantee optimal results under all traffic scenarios. Thus, ideal dynamic route guidance methods should have a highly adaptive learning ability under diverse traffic scenarios so as to have extensive improvement capabilities for different traffic scenarios. In this study, an A* trajectory rejection method based on multi-agent reinforcement learning (A*R²) is proposed; the method integrates both system and user perspectives to mitigate traffic congestion and reduce travel time (TT) and travel distance (TD). First, owing to its adaptive learning ability, the A*R² can comprehensively analyze the traffic conditions for different traffic scenarios and intelligently evaluate the road congestion index from a system perspective. Then, the A*R² determines the routes for all vehicles from user perspective according to the road network congestion index. An extensive set of simulation experiments reveal that, under various traffic scenarios, the A*R² can rely on its adaptive learning ability to achieve better traffic efficiency. Moreover, even in cases where many drivers are not fully compliant with the route guidance, the traffic efficiency can still be improved significantly by A*R².
Yu J, Mo D, Xie N, et al., 2021, Exploring multi-homing behavior of ride-sourcing drivers via real-world multiple platforms data, TRANSPORTATION RESEARCH PART F-TRAFFIC PSYCHOLOGY AND BEHAVIOUR, Vol: 80, Pages: 61-78, ISSN: 1369-8478
- Author Web Link
- Cite
- Citations: 4
Qian G, Guo M, Zhang L, et al., 2021, Traffic scheduling and control in fully connected and automated networks, TRANSPORTATION RESEARCH PART C-EMERGING TECHNOLOGIES, Vol: 126, ISSN: 0968-090X
- Author Web Link
- Cite
- Citations: 10
Wang Y, Yu X, Zhang S, et al., 2021, Freeway Traffic Control in Presence of Capacity Drop, IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS, Vol: 22, Pages: 1497-1516, ISSN: 1524-9050
- Author Web Link
- Cite
- Citations: 7
Liu X, Zheng R, Wang H, et al., 2021, A Knowledge Management Framework for Vehicle Hazard Analysis, Pages: 165-169
Hazard analysis is an essential activity in the life-cycle of vehicle development. Automotive firms use various hazard analysis techniques to identify hazards and to perform safety analysis. However, these methods are time-consuming and requires significant effort for brainstorming and discussion. Increasing the efficiency of the hazard analysis process is essential to enhance an automotive firm's competitive advantage. Since a large amount of data and the reuse of the original analysis case can undoubtedly improve analysis efficiency, knowledge management is a feasible solution to enhance hazard analysis efficacy. This paper proposes a knowledge management framework to automatically capture and mine knowledge resources from accumulated data. This framework can assist engineers to identify the items in the hazard analysis process and increase their productivity.
Ma Y, Wang L, Wang Y, et al., 2021, Developing Smart Lane-changing Strategies for CAVs on Freeways based on MOBIL and Reinforcement Learning, IEEE Intelligent Transportation Systems Conference (ITSC), Publisher: IEEE, Pages: 2027-2033, ISSN: 2153-0009
Hu Y, Wang Y, Jin X, et al., 2021, Urban Eco-driving of Connected and Automated Vehicles in Traffic-Mixed and Power-heterogeneous Conditions, IEEE Intelligent Transportation Systems Conference (ITSC), Publisher: IEEE, Pages: 873-878, ISSN: 2153-0009
Zhou Q, Mohammadi R, Zhao W, et al., 2021, Queue Profile Identification at Signalized Intersections with High-Resolution Data from Drones, 7th International Conference on Models and Technologies for Intelligent Transportation Systems (MT-ITS), Publisher: IEEE
- Author Web Link
- Cite
- Citations: 1
Li J, Zhang K, Shen L, et al., 2021, A Domain Adaptation Framework for Short-term Traffic Prediction, IEEE Intelligent Transportation Systems Conference (ITSC), Publisher: IEEE, Pages: 3564-3569, ISSN: 2153-0009
Taleongpong P, Hu S, Jiang Z, et al., 2020, Machine learning techniques to predict reactionary delays and other associated key performance indicators on British railway network, Journal of Intelligent Transportation Systems: technology, planning, and operations, Vol: 2020, Pages: 1-19, ISSN: 1547-2450
Reactionary delays that propagate from a primary source throughout train journeys are an immediate concern for British railway systems. Complex non-linear interactions between various spatiotemporal variables govern the propagation of these delays which can avalanche throughout railway network causing further severe disruptions. This paper introduces several machine learning (ML) techniques alongside data mining processes to create a framework that predicts key performance indicators (KPIs), reactionary arrival delay, reactionary departure delay, dwell time and travel time. The frameworks in this paper provide greater accuracy in predicting KPIs through state-of-the-art ML models compared to existing delay prediction systems. Further discussion on the improvements, applicability and scalability of this framework are also provided in this paper.
Li J, Guo F, Wang Y, et al., 2020, Short-term traffic prediction with deep neural networks and adaptive transfer learning, 23rd International Conference on Intelligent Transportation Systems (ITSC), Publisher: IEEE, Pages: 1-6
A key problem in short-term traffic prediction is the prevailing data missing scenarios across the entire traffic network. To address this challenge, a transfer learning framework is currently used in the literature, which could improve the prediction accuracy on the target link that suffers severe data missing problems by using information from source links with sufficient historical data. However, one of the limitations in these transfer-learning based models is their high dependency on the consistency between datasets and the complex data selection process, which brings heavy computation burden and human efforts. In this paper, we propose an adaptive transfer learning method in short-term traffic flow prediction model to alleviate the complex data selection process. Specifically, a self-adaptive neural network with a novel domain adaptation loss is developed. The domain adaptation loss is able to calculate the distance between the source data and the corresponding target data in each training batch, which can help the network to adaptively filter inconsistent source data and learn target link related information in each training batch. The Maximum Mean Discrepancy (MMD) measurement, which has been fully validated and applied in transfer learning research, is used in combination with the Gaussian kernel to measure the distance between datasets in each training batch. A series of experiments are designed and conducted using 15-minute interval traffic flow data from the Highways England, UK. The results have demonstrated that the proposed adaptive transfer learning method is less affected by the inconsistency between datasets and provides more accurate short-term traffic flow prediction.
Wu C, Hu S, Lee C-H, et al., 2020, Multi-platform data collection for public service with Pay-by-Data, MULTIMEDIA TOOLS AND APPLICATIONS, Vol: 79, Pages: 33503-33518, ISSN: 1380-7501
Liu W, van Hooff T, An Y, et al., 2020, Modeling transient particle transport in transient indoor airflow by fast fluid dynamics with the Markov chain method, Building and Environment, Vol: 186, Pages: 1-11, ISSN: 0360-1323
It is crucial to accurately and efficiently predict transient particle transport in indoor environments to improve air distribution design and reduce health risks. For steady-state indoor airflow, fast fluid dynamics (FFD) + Markov chain model increased the calculation speed by around seven times compared to computational fluid dynamics (CFD) + Eulerian model and CFD + Lagrangian model, while achieving the same level of accuracy. However, the indoor airflow could be transient, if there were human behaviors involved like coughing or sneezing and air was supplied periodically. Therefore, this study developed an FFD + Markov chain model solver for predicting transient particle transport in transient indoor airflow. This investigation used two cases, transient particle transport in a ventilated two-zone chamber and a chamber with periodic air supplies, for validation. Case 1 had experimental data for validation and the results showed that the predicted particle concentration by FFD + Markov chain model matched well with the experimental data. Besides, it had similar accuracy as the CFD + Eulerian model. In the second case, the prediction by large eddy simulation (LES) was used for validating the FFD. The simulated particle concentrations by the Markov chain model and the Eulerian model were similar. The simulated particle concentrations by the Markov chain model and the Eulerian model were similar. The computational time of the FFD + Markov chain model was 7.8 times less than that of the CFD + Eulerian model.
This data is extracted from the Web of Science and reproduced under a licence from Thomson Reuters. You may not copy or re-distribute this data in whole or in part without the written consent of the Science business of Thomson Reuters.