Journal of Intelligent Transportation Systems最新文献_第5页

DAnoScenE: a driving anomaly scenario extraction framework for autonomous vehicles in urban streets DAnoScenE：城市街道自动驾驶车辆异常驾驶场景提取框架

IF 2.8 3区工程技术 Q3 TRANSPORTATION

Journal of Intelligent Transportation Systems

Pub Date : 2025-01-02 Epub Date: 2023-12-08 DOI: 10.1080/15472450.2023.2291680

Yuening Hu , Dan Zhao , Ying Wang , Guangming Zhao

Autonomous vehicles (AVs) hold great potential to improve traffic safety. However, urban streets present a dynamic environment where unforeseen and complex scenarios can arise. The establishment of a systematic framework to extract a variety of vehicle driving scenarios could empower AVs to learn from and effectively navigate various situations. This study introduces a driving anomaly scenario extraction (DAnoScenE) framework tailored for AVs operating in urban street settings. The Waymo Open Motion Dataset (WOMD) is used to showcase the framework’s capability to capture an extensive range of realistic driving anomaly scenarios. The central process involves the detection and labeling of driving anomalies. To avoid the erroneous detected and labeled driving anomalies arising from issues such as outliers and noise within vehicle track data, a two-step approach is introduced to analyze and rectify vehicle movement parameters in raw data. To comprehend these driving anomalies and their associated scenarios, manual labeling identifies causative factors of scenarios such as traffic signals and behaviors of other agents, forming three scenario groups: Signal Interaction, Agent Interaction, and Other. A multimodal model is developed to classify scenario groups, complemented by a segmentation process that further divides groups into specific scenarios based on simple conditions. The results show that recognition accuracy for driving anomaly scenario groups achieved 98.4%, and the scenario segmentation method achieved 100% accuracy by simple conditions. The proposed framework provides valuable support for the advancement of autonomous driving algorithms and comprehensive AV testing, with a specific emphasis on navigating abnormal driving environments.

自动驾驶汽车（AV）在改善交通安全方面具有巨大潜力。然而，城市街道是一个动态环境，可能会出现不可预见的复杂情况。建立...

{"title":"DAnoScenE: a driving anomaly scenario extraction framework for autonomous vehicles in urban streets","authors":"Yuening Hu , Dan Zhao , Ying Wang , Guangming Zhao","doi":"10.1080/15472450.2023.2291680","DOIUrl":"10.1080/15472450.2023.2291680","url":null,"abstract":"<div><div>Autonomous vehicles (AVs) hold great potential to improve traffic safety. However, urban streets present a dynamic environment where unforeseen and complex scenarios can arise. The establishment of a systematic framework to extract a variety of vehicle driving scenarios could empower AVs to learn from and effectively navigate various situations. This study introduces a driving anomaly scenario extraction (DAnoScenE) framework tailored for AVs operating in urban street settings. The Waymo Open Motion Dataset (WOMD) is used to showcase the framework’s capability to capture an extensive range of realistic driving anomaly scenarios. The central process involves the detection and labeling of driving anomalies. To avoid the erroneous detected and labeled driving anomalies arising from issues such as outliers and noise within vehicle track data, a two-step approach is introduced to analyze and rectify vehicle movement parameters in raw data. To comprehend these driving anomalies and their associated scenarios, manual labeling identifies causative factors of scenarios such as traffic signals and behaviors of other agents, forming three scenario groups: Signal Interaction, Agent Interaction, and Other. A multimodal model is developed to classify scenario groups, complemented by a segmentation process that further divides groups into specific scenarios based on simple conditions. The results show that recognition accuracy for driving anomaly scenario groups achieved 98.4%, and the scenario segmentation method achieved 100% accuracy by simple conditions. The proposed framework provides valuable support for the advancement of autonomous driving algorithms and comprehensive AV testing, with a specific emphasis on navigating abnormal driving environments.</div></div>","PeriodicalId":54792,"journal":{"name":"Journal of Intelligent Transportation Systems","volume":"29 1","pages":"Pages 32-52"},"PeriodicalIF":2.8,"publicationDate":"2025-01-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138689862","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Modified Gipps model: a collision-free car following model 改良吉普斯模型：无碰撞汽车跟随模型

IF 2.8 3区工程技术 Q3 TRANSPORTATION

Journal of Intelligent Transportation Systems

Pub Date : 2025-01-02 Epub Date: 2023-12-15 DOI: 10.1080/15472450.2023.2289149

Dhwani Shah , Chris Lee , Yong Hoon Kim

Car following (CF) models are used in microscopic traffic simulation tools to help assess the effects of a new road design or to assess the effect of change in traffic flow. In 1981, Gipps developed a collision avoidance CF model using the Newtonian laws of motion to describe the motion of each vehicle in a stream of traffic. It is one of the most widely used CF models in both research and practice. Although it is claimed that the Gipps model produces collision-free results, the model produces a collision when the intention of the following vehicle is to brake harder than the perceived deceleration of lead vehicle. For the ease of simulations, a traffic simulation tool is expected to not show unrealistic crashes. This study was carried out to make the Gipps model collision-free in all conditions. It first highlights the conditions where the original Gipps model produces a collision. Then the study derives an equation for a collision-free Gipps CF model. This modified Gipps CF model produces collision-free results that always maintain a safe spacing with the lead vehicle.

微观交通模拟工具中使用的汽车跟随（CF）模型可帮助评估新道路设计的效果或交通流量变化的影响。1981 年，吉普斯开发了汽车跟随模型。

引用次数: 0

Generative adversarial network for car following trajectory generation and anomaly detection 用于生成汽车行驶轨迹和异常检测的生成对抗网络

IF 2.8 3区工程技术 Q3 TRANSPORTATION

Journal of Intelligent Transportation Systems

Pub Date : 2025-01-02 Epub Date: 2024-01-06 DOI: 10.1080/15472450.2023.2301691

Haotian Shi , Shuoxuan Dong , Yuankai Wu , Qinghui Nie , Yang Zhou , Bin Ran

Car-following trajectory generation and anomaly detection are critical functions in the sensing module of an automated vehicle. However, developing models that capture realistic trajectory data distribution and detect anomalous driving behaviors could be challenging. This paper proposes ‘TrajGAN’, an unsupervised learning approach based on the Generative Adversarial Network (GAN) to exploit vehicle car following trajectory data for generation and anomaly detection. The proposed TrajGAN consists of two modules, an encoder-decoder Long Short-Term Memory (LSTM)-based generator and an LSTM-multilayer perceptron (MLP) based discriminator, whose former component is used to generate vehicular car following trajectories and the latter one is for trajectory anomaly detection. By letting these two modules game with each other in training, we can simultaneously achieve robust trajectory generators and anomaly detectors. Trained with the Next Generation Simulation (NGSIM) dataset, TrajGAN can generate realistic trajectories with a similar distribution of training data and identify a manifold of anomalous trajectories based on an anomaly scoring scheme. Simulation results indicate that the proposed approach is efficient in reproducing artificial trajectories and identifying anomalous driving behaviors.

汽车行驶轨迹生成和异常检测是自动驾驶汽车传感模块的关键功能。然而，开发能捕捉真实轨迹数据的模型并不容易。

{"title":"Generative adversarial network for car following trajectory generation and anomaly detection","authors":"Haotian Shi , Shuoxuan Dong , Yuankai Wu , Qinghui Nie , Yang Zhou , Bin Ran","doi":"10.1080/15472450.2023.2301691","DOIUrl":"10.1080/15472450.2023.2301691","url":null,"abstract":"<div><div>Car-following trajectory generation and anomaly detection are critical functions in the sensing module of an automated vehicle. However, developing models that capture realistic trajectory data distribution and detect anomalous driving behaviors could be challenging. This paper proposes ‘TrajGAN’, an unsupervised learning approach based on the Generative Adversarial Network (GAN) to exploit vehicle car following trajectory data for generation and anomaly detection. The proposed TrajGAN consists of two modules, an encoder-decoder Long Short-Term Memory (LSTM)-based generator and an LSTM-multilayer perceptron (MLP) based discriminator, whose former component is used to generate vehicular car following trajectories and the latter one is for trajectory anomaly detection. By letting these two modules game with each other in training, we can simultaneously achieve robust trajectory generators and anomaly detectors. Trained with the Next Generation Simulation (NGSIM) dataset, TrajGAN can generate realistic trajectories with a similar distribution of training data and identify a manifold of anomalous trajectories based on an anomaly scoring scheme. Simulation results indicate that the proposed approach is efficient in reproducing artificial trajectories and identifying anomalous driving behaviors.</div></div>","PeriodicalId":54792,"journal":{"name":"Journal of Intelligent Transportation Systems","volume":"29 1","pages":"Pages 53-66"},"PeriodicalIF":2.8,"publicationDate":"2025-01-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139766121","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Data-driven transfer learning framework for estimating on-ramp and off-ramp traffic flows 用于估算匝道和非匝道交通流量的数据驱动迁移学习框架

IF 2.8 3区工程技术 Q3 TRANSPORTATION

Journal of Intelligent Transportation Systems

Pub Date : 2025-01-02 Epub Date: 2024-01-03 DOI: 10.1080/15472450.2023.2301696

Xiaobo Ma , Abolfazl Karimpour , Yao-Jan Wu

To develop the most appropriate control strategy and monitor, maintain, and evaluate the traffic performance of the freeway weaving areas, state and local Departments of Transportation need to have access to traffic flows at each pair of on-ramp and off-ramp. However, ramp flows are not always readily available to transportation agencies, and little effort has been made to estimate these missing traffic flows in locations where no physical sensors are installed. To bridge this research gap, a data-driven framework is proposed that can accurately estimate the missing ramp flows by solely using data collected from loop detectors on freeway mainlines. The proposed framework employs a transfer learning model. The transfer learning model relaxes the assumption that the underlying data distributions of the source and target domains must be the same. Therefore, the proposed framework can guarantee high-accuracy estimation of on-ramp and off-ramp flows on freeways with different traffic patterns, distributions, and characteristics. Based on the experimental results, the flow estimation mean absolute errors range between 23.90 veh/h to 40.85 veh/h for on-ramps and 31.58 veh/h to 45.31 veh/h for off-ramps; the flow estimation root mean square errors range between 34.55 veh/h to 57.77 veh/h for on-ramps, and 41.75 veh/h to 58.80 veh/h for off-ramps. Further, the comparison analysis shows that the proposed framework outperforms other conventional machine learning models. The estimated ramp flows based on the proposed method can help transportation agencies to enhance the operations of their ramp control strategies for locations where physical sensors are not installed.

为了制定最合适的控制策略，监测、维护和评估高速公路交织区的交通性能，各州和地方交通部门需要...

{"title":"Data-driven transfer learning framework for estimating on-ramp and off-ramp traffic flows","authors":"Xiaobo Ma , Abolfazl Karimpour , Yao-Jan Wu","doi":"10.1080/15472450.2023.2301696","DOIUrl":"10.1080/15472450.2023.2301696","url":null,"abstract":"<div><div>To develop the most appropriate control strategy and monitor, maintain, and evaluate the traffic performance of the freeway weaving areas, state and local Departments of Transportation need to have access to traffic flows at each pair of on-ramp and off-ramp. However, ramp flows are not always readily available to transportation agencies, and little effort has been made to estimate these missing traffic flows in locations where no physical sensors are installed. To bridge this research gap, a data-driven framework is proposed that can accurately estimate the missing ramp flows by solely using data collected from loop detectors on freeway mainlines. The proposed framework employs a transfer learning model. The transfer learning model relaxes the assumption that the underlying data distributions of the source and target domains must be the same. Therefore, the proposed framework can guarantee high-accuracy estimation of on-ramp and off-ramp flows on freeways with different traffic patterns, distributions, and characteristics. Based on the experimental results, the flow estimation mean absolute errors range between 23.90 veh/h to 40.85 veh/h for on-ramps and 31.58 veh/h to 45.31 veh/h for off-ramps; the flow estimation root mean square errors range between 34.55 veh/h to 57.77 veh/h for on-ramps, and 41.75 veh/h to 58.80 veh/h for off-ramps. Further, the comparison analysis shows that the proposed framework outperforms other conventional machine learning models. The estimated ramp flows based on the proposed method can help transportation agencies to enhance the operations of their ramp control strategies for locations where physical sensors are not installed.</div></div>","PeriodicalId":54792,"journal":{"name":"Journal of Intelligent Transportation Systems","volume":"29 1","pages":"Pages 67-80"},"PeriodicalIF":2.8,"publicationDate":"2025-01-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139470605","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Smart Mobility in Smart Cities: Emerging challenges, recent advances and future directions 智慧城市中的智慧交通：新挑战、最新进展和未来方向

IF 2.8 3区工程技术 Q3 TRANSPORTATION

Journal of Intelligent Transportation Systems

Pub Date : 2025-01-02 Epub Date: 2023-08-08 DOI: 10.1080/15472450.2023.2245750

Soumia Goumiri , Saïd Yahiaoui , Soufiene Djahel

The world is witnessing a vivid race toward developing advanced solutions to enable smart, fast, affordable and environment friendly mobility for Smart Cities inhabitants. This led to the emergence of the Smart Mobility concept, attracting significant attention from major actors in the mobility sector including policy makers and traffic authorities. Therefore, this survey paper presents an overview of Smart Mobility and discusses the main challenges associated with its key building blocks, parking and traffic management, traffic routing in addition to emissions and road safety implications. Then, the most important works that attempted to address these challenges are presented, and their strengths and limitations are analyzed. Finally, the lessons learned from this study and the most promising future directions to tackle these challenges are presented.

世界正在见证一场生动的竞赛，即开发先进的解决方案，为智慧城市居民提供智能、快速、经济、环保的出行方式。这导致了智能移动概念的出现，吸引了包括政策制定者和交通部门在内的移动领域主要参与者的极大关注。因此，本调查报告概述了智能移动出行，并讨论了与其关键构建模块、停车和交通管理、交通路线以及排放和道路安全影响相关的主要挑战。然后，介绍了试图解决这些挑战的最重要的作品，并分析了它们的优势和局限性。最后，提出了从本研究中得到的经验教训以及应对这些挑战的最有希望的未来方向。

引用次数: 0

Deep learning-based pedestrian trajectory prediction and risk assessment at signalized intersections using trajectory data captured through roadside LiDAR 基于深度学习的行人轨迹预测和风险评估，使用路边激光雷达捕获的轨迹数据

IF 2.8 3区工程技术 Q3 TRANSPORTATION

Journal of Intelligent Transportation Systems

Pub Date : 2024-11-01 Epub Date: 2023-05-02 DOI: 10.1080/15472450.2023.2209912

Shanglian Zhou , Hao Xu , Guohui Zhang , Tianwei Ma , Yin Yang

In recent years, rapid advancements in the Autonomous Vehicles (AVs) industry have greatly motivated the research and development in pedestrian trajectory prediction and risk assessment. One of the critical requirements for AVs is to predict the future trajectories of pedestrians and provide collision warnings in an accurate and prompt manner. Nevertheless, accurate prediction of pedestrian trajectories remains a technical challenge, mainly caused by the heterogeneity of pedestrian crossing behavior and uncertainties in vehicle-pedestrian interactions. This paper proposes a deep learning-based method for pedestrian trajectory prediction and risk assessment, using trajectory data extracted from roadside LiDAR data and corresponding signal phasing information at MLK and Georgia Avenue in Chattanooga, TN. Meanwhile, a set of criteria referred to as the risk factor is established to quantitatively evaluate the risk of the pedestrian crossing behavior, which also serves as a learnable feature. A Long Short-Term Memory (LSTM) network is proposed, which takes the following data as the input: the pedestrian trajectory data, signal phasing data, and risk factors from the past 10 steps. Meanwhile, the network predicts the pedestrian trajectory and risk factor at the future time step. In the experimental study, the root-mean-square errors between the predicted and ground truth x and y coordinates are 0.225 meters and 0.377 meters, respectively, and the F1 score value for the risk factor is 99.6%, demonstrating the efficacy of the proposed LSTM-based methodology on pedestrian trajectory prediction and risk assessment.

近年来，自动驾驶汽车行业的快速发展极大地推动了行人轨迹预测和风险评估的研究与发展。自动驾驶汽车的关键要求之一是预测行人的未来轨迹，并以准确和及时的方式提供碰撞警告。然而，行人轨迹的准确预测仍然是一个技术挑战，主要是由于行人过马路行为的异质性和车-人相互作用的不确定性。本文提出了一种基于深度学习的行人轨迹预测和风险评估方法，该方法利用美国田纳西州查塔努加（Chattanooga）的MLK和Georgia Avenue的路边LiDAR数据提取的轨迹数据和相应的信号相位信息，同时建立了一套称为风险因子的标准，定量评估行人过马路行为的风险，这也是一种可学习的特征。提出了一种长短期记忆（LSTM）网络，该网络以以下数据作为输入：行人轨迹数据、信号相位数据和过去10步的风险因素。同时，该网络预测未来时间步长的行人轨迹和危险因素。在实验研究中，预测值与真实值x、y坐标的均方根误差分别为0.225米和0.377米，风险因子F1得分值为99.6%，证明了基于lstm的行人轨迹预测和风险评估方法的有效性。

{"title":"Deep learning-based pedestrian trajectory prediction and risk assessment at signalized intersections using trajectory data captured through roadside LiDAR","authors":"Shanglian Zhou , Hao Xu , Guohui Zhang , Tianwei Ma , Yin Yang","doi":"10.1080/15472450.2023.2209912","DOIUrl":"10.1080/15472450.2023.2209912","url":null,"abstract":"<div><div>In recent years, rapid advancements in the Autonomous Vehicles (AVs) industry have greatly motivated the research and development in pedestrian trajectory prediction and risk assessment. One of the critical requirements for AVs is to predict the future trajectories of pedestrians and provide collision warnings in an accurate and prompt manner. Nevertheless, accurate prediction of pedestrian trajectories remains a technical challenge, mainly caused by the heterogeneity of pedestrian crossing behavior and uncertainties in vehicle-pedestrian interactions. This paper proposes a deep learning-based method for pedestrian trajectory prediction and risk assessment, using trajectory data extracted from roadside LiDAR data and corresponding signal phasing information at MLK and Georgia Avenue in Chattanooga, TN. Meanwhile, a set of criteria referred to as the risk factor is established to quantitatively evaluate the risk of the pedestrian crossing behavior, which also serves as a learnable feature. A Long Short-Term Memory (LSTM) network is proposed, which takes the following data as the input: the pedestrian trajectory data, signal phasing data, and risk factors from the past 10 steps. Meanwhile, the network predicts the pedestrian trajectory and risk factor at the future time step. In the experimental study, the root-mean-square errors between the predicted and ground truth <em>x</em> and <em>y</em> coordinates are 0.225 meters and 0.377 meters, respectively, and the F1 score value for the risk factor is 99.6%, demonstrating the efficacy of the proposed LSTM-based methodology on pedestrian trajectory prediction and risk assessment.</div></div>","PeriodicalId":54792,"journal":{"name":"Journal of Intelligent Transportation Systems","volume":"28 6","pages":"Pages 793-805"},"PeriodicalIF":2.8,"publicationDate":"2024-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"88124815","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Inferring the number of vehicles between trajectory-observed vehicles 推断轨迹观测车辆之间的车辆数量

IF 2.8 3区工程技术 Q3 TRANSPORTATION

Journal of Intelligent Transportation Systems

Pub Date : 2024-11-01 Epub Date: 2023-09-01 DOI: 10.1080/15472450.2023.2227940

Zhiyong Wen , Xiaoxiong Weng

Traffic perception is the foundation of intelligent roads, and how to accurately perceive traffic has become a central issue for researchers. With the application of Vehicle-to-Everything communication technology, vehicle IDs, locations, velocities, and accelerations can be obtained by the Roadside Unit (RSU), i.e., trajectory-observed vehicles for the road. Inferring the number of vehicles between trajectory-observed vehicles can make traffic perception more accurate, with which the traffic can be sensed on the whole road. Thus, in the case of mixed traffic flow, a Real-Time Prediction Model was proposed, which is a novel model containing four modules: prior experience of the space headway, linear distribution of velocity and acceleration, identification of traffic shockwave, and filter. The inferred result was calculated in real time. During the test, we used US-101 lane-1 data of the Next Generation Simulation dataset and trajectory-observed vehicles with stochastic distribution for 20% penetration. The length of the study area on the US-101 highway was approximately 2100 feet, which was similar to the communication area of a single RSU. During the evaluation of the model accuracy with the real-world datasets, the error of the inferred vehicle numbers in the study area could be limited to ±5 vehicles almost. Results show that it is feasible to infer the number of vehicles between trajectory-observed vehicles. The model compensates for the shortcomings of traditional models (based on inductive loop, camera, or radar), thus providing a novel method for the traffic perception of intelligent roads.

交通感知是智能道路的基础，如何准确感知交通已成为研究人员关注的核心问题。随着车对一切通信技术的应用，路边单元（RSU），即道路上的轨迹观测车辆，可以获得车辆的id、位置、速度和加速度。推断轨迹观测车辆之间的车辆数量可以提高交通感知的准确性，从而可以感知整个道路的交通状况。为此，提出了混合交通流实时预测模型，该模型包含车头距先验经验、速度和加速度线性分布、交通冲击波识别和滤波四个模块。对推断结果进行实时计算。在测试过程中，我们使用了US-101的下一代模拟数据集的1号车道数据和随机分布的轨迹观察车辆，渗透率为20%。US-101高速公路上的研究区域长度约为2100英尺，与单个RSU的通信区域相似。在使用实际数据集评估模型精度时，研究区域推断的车辆数量误差几乎可以限制在±5辆。结果表明，在轨迹观测车辆之间推断车辆数量是可行的。该模型弥补了传统模型（基于感应回路、摄像头或雷达）的不足，为智能道路的交通感知提供了一种新的方法。

{"title":"Inferring the number of vehicles between trajectory-observed vehicles","authors":"Zhiyong Wen , Xiaoxiong Weng","doi":"10.1080/15472450.2023.2227940","DOIUrl":"10.1080/15472450.2023.2227940","url":null,"abstract":"<div><div>Traffic perception is the foundation of intelligent roads, and how to accurately perceive traffic has become a central issue for researchers. With the application of Vehicle-to-Everything communication technology, vehicle IDs, locations, velocities, and accelerations can be obtained by the Roadside Unit (RSU), i.e., trajectory-observed vehicles for the road. Inferring the number of vehicles between trajectory-observed vehicles can make traffic perception more accurate, with which the traffic can be sensed on the whole road. Thus, in the case of mixed traffic flow, a Real-Time Prediction Model was proposed, which is a novel model containing four modules: prior experience of the space headway, linear distribution of velocity and acceleration, identification of traffic shockwave, and filter. The inferred result was calculated in real time. During the test, we used US-101 lane-1 data of the Next Generation Simulation dataset and trajectory-observed vehicles with stochastic distribution for 20% penetration. The length of the study area on the US-101 highway was approximately 2100 feet, which was similar to the communication area of a single RSU. During the evaluation of the model accuracy with the real-world datasets, the error of the inferred vehicle numbers in the study area could be limited to ±5 vehicles almost. Results show that it is feasible to infer the number of vehicles between trajectory-observed vehicles. The model compensates for the shortcomings of traditional models (based on inductive loop, camera, or radar), thus providing a novel method for the traffic perception of intelligent roads.</div></div>","PeriodicalId":54792,"journal":{"name":"Journal of Intelligent Transportation Systems","volume":"28 6","pages":"Pages 816-829"},"PeriodicalIF":2.8,"publicationDate":"2024-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74968740","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

How spatial features affect urban rail transit prediction accuracy: a deep learning based passenger flow prediction method 空间特征如何影响城市轨道交通预测精度：基于深度学习的客流预测方法

IF 2.8 3区工程技术 Q3 TRANSPORTATION

Journal of Intelligent Transportation Systems

Pub Date : 2024-11-01 Epub Date: 2023-11-06 DOI: 10.1080/15472450.2023.2279633

Shuang Li , Xiaoxi Liang , Meina Zheng , Junlan Chen , Ting Chen , Xiucheng Guo

Urban rail transit is an integral part of public transit, and has been extensive built in China. Previous studies have proved that the spatial features are closely related to rail transit ridership, considering a fundamental role of short-term passenger flow forecast in the urban rail operation, it is meaningful to explore how these factors affect the prediction accuracy. This study aims to find a way to improve prediction accuracy by considering spatial features of stations based on deep learning. Therefore, a CNN-LSTM model capturing the spatial and temporal features was applied and Suzhou (China) was choosing as a case study to explore the influence of three spatial features, namely relative position, location, and land use, on the prediction accuracy. The predict model used can extract spatiotemporal features and accurately predict the citywide stations, and the results show that, for the relative position, the inbound and outbound flow prediction errors of transfer stations and middle stations are the lowest, respectively. As for locational features, the more distant the station is from the city center, the more accurate the results are. For stations where land use is dominated by work and living services, the predictions are more accurate. The error rate is higher for stations whose services are mainly tourism, transportation, and leisure services. This study’s results can help operators predict the short-term passenger flow of target stations based on different demands and optimize their services on this basis.

城市轨道交通是公共交通的重要组成部分，在中国得到了广泛的建设。已有研究证明，空间特征与轨道交通客流量密切相关，考虑到短期客流预测在城市轨道交通运营中的基础性作用，探讨这些因素对预测精度的影响具有重要意义。本研究旨在寻找一种基于深度学习的方法，通过考虑台站的空间特征来提高预测精度。因此，采用CNN-LSTM模型捕捉时空特征，并以中国苏州为例，探讨相对位置、地理位置和土地利用三个空间特征对预测精度的影响。所建立的预测模型能够提取时空特征，准确预测全市范围内的站点，结果表明：对于相对位置，中转站的进站流量预测误差最小，中转站的出站流量预测误差最小；在区位特征方面，站点离市中心越远，结果越准确。对于那些土地使用以工作和生活服务为主的车站，预测更为准确。以旅游、交通和休闲服务为主的站点错误率较高。研究结果可以帮助运营商根据不同需求预测目标站点的短期客流，并在此基础上优化服务。

{"title":"How spatial features affect urban rail transit prediction accuracy: a deep learning based passenger flow prediction method","authors":"Shuang Li , Xiaoxi Liang , Meina Zheng , Junlan Chen , Ting Chen , Xiucheng Guo","doi":"10.1080/15472450.2023.2279633","DOIUrl":"10.1080/15472450.2023.2279633","url":null,"abstract":"<div><div>Urban rail transit is an integral part of public transit, and has been extensive built in China. Previous studies have proved that the spatial features are closely related to rail transit ridership, considering a fundamental role of short-term passenger flow forecast in the urban rail operation, it is meaningful to explore how these factors affect the prediction accuracy. This study aims to find a way to improve prediction accuracy by considering spatial features of stations based on deep learning. Therefore, a CNN-LSTM model capturing the spatial and temporal features was applied and Suzhou (China) was choosing as a case study to explore the influence of three spatial features, namely relative position, location, and land use, on the prediction accuracy. The predict model used can extract spatiotemporal features and accurately predict the citywide stations, and the results show that, for the relative position, the inbound and outbound flow prediction errors of transfer stations and middle stations are the lowest, respectively. As for locational features, the more distant the station is from the city center, the more accurate the results are. For stations where land use is dominated by work and living services, the predictions are more accurate. The error rate is higher for stations whose services are mainly tourism, transportation, and leisure services. This study’s results can help operators predict the short-term passenger flow of target stations based on different demands and optimize their services on this basis.</div></div>","PeriodicalId":54792,"journal":{"name":"Journal of Intelligent Transportation Systems","volume":"28 6","pages":"Pages 1032-1043"},"PeriodicalIF":2.8,"publicationDate":"2024-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135036902","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Adaptive green split optimization for traffic control with low penetration rate trajectory data 低穿透率轨迹数据下交通控制的自适应绿裂优化

IF 2.8 3区工程技术 Q3 TRANSPORTATION

Journal of Intelligent Transportation Systems

Pub Date : 2024-11-01 Epub Date: 2023-08-29 DOI: 10.1080/15472450.2023.2227959

Zihao Wang , Roger Lloret-Batlle , Jianfeng Zheng , Henry X. Liu

Adaptive traffic signal control systems often rely on expensive physical detection infrastructure. However, with the advent of widespread trajectory data, it is now possible to implement adaptive control entirely avoiding such costs. We present two simple adaptive control policies which only require sample delay and number of stops, with the goal to mitigate the presence of oversaturation. The simplicity stems from the necessity of controlling under any trajectory penetration rate. The two policies differ on the possibilities of the control infrastructure to be implemented. The first one minimizes oversaturation by deviating from a reference pre-timed signal plan. This signal plan can be an existing one or an estimated one from aggregating trajectory data. The second policy creates first a set of green split plans to be then selected by a control logic. This second policy is intended to be used in SCATS-like systems where signal plans are limited to a pre-defined discrete set. We propose a plan selection logics or alternatively, the original plan selection policy can be used as well. Both policies are tested in the field, achieving a significant reduction in delay, oversaturation and spillover ratios. Lastly, we test an application of this policy as an enhancement of SCATS systems in the presence of malfunctioning physical detectors.

自适应交通信号控制系统通常依赖于昂贵的物理检测基础设施。然而，随着广泛的轨迹数据的出现，现在可以实现完全避免此类成本的自适应控制。我们提出了两种简单的自适应控制策略，它们只需要采样延迟和停止次数，目的是减轻过饱和的存在。之所以简单，是因为需要在任意弹道突防速率下进行控制。这两种策略在实现控制基础设施的可能性上有所不同。第一种是通过偏离参考的预定时信号计划来最小化过饱和。该信号平面可以是现有的信号平面，也可以是综合轨迹数据估计出来的信号平面。第二个策略首先创建一组绿色分割计划，然后由控制逻辑选择。第二个策略旨在用于类似scats的系统，其中信号计划仅限于预定义的离散集。我们提出了一个计划选择逻辑，或者，也可以使用原始的计划选择策略。这两项政策都在实地进行了测试，显著降低了延迟、过饱和和溢出比率。最后，我们测试了该策略的应用，作为SCATS系统在存在故障物理探测器的情况下的增强。

{"title":"Adaptive green split optimization for traffic control with low penetration rate trajectory data","authors":"Zihao Wang , Roger Lloret-Batlle , Jianfeng Zheng , Henry X. Liu","doi":"10.1080/15472450.2023.2227959","DOIUrl":"10.1080/15472450.2023.2227959","url":null,"abstract":"<div><div>Adaptive traffic signal control systems often rely on expensive physical detection infrastructure. However, with the advent of widespread trajectory data, it is now possible to implement adaptive control entirely avoiding such costs. We present two simple adaptive control policies which only require sample delay and number of stops, with the goal to mitigate the presence of oversaturation. The simplicity stems from the necessity of controlling under any trajectory penetration rate. The two policies differ on the possibilities of the control infrastructure to be implemented. The first one minimizes oversaturation by deviating from a reference pre-timed signal plan. This signal plan can be an existing one or an estimated one from aggregating trajectory data. The second policy creates first a set of green split plans to be then selected by a control logic. This second policy is intended to be used in SCATS-like systems where signal plans are limited to a pre-defined discrete set. We propose a plan selection logics or alternatively, the original plan selection policy can be used as well. Both policies are tested in the field, achieving a significant reduction in delay, oversaturation and spillover ratios. Lastly, we test an application of this policy as an enhancement of SCATS systems in the presence of malfunctioning physical detectors.</div></div>","PeriodicalId":54792,"journal":{"name":"Journal of Intelligent Transportation Systems","volume":"28 6","pages":"Pages 830-845"},"PeriodicalIF":2.8,"publicationDate":"2024-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73973599","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Adaptive graph convolutional network-based short-term passenger flow prediction for metro 基于自适应图卷积网络的地铁短期客流预测

IF 2.8 3区工程技术 Q3 TRANSPORTATION

Journal of Intelligent Transportation Systems

Pub Date : 2024-11-01 Epub Date: 2023-05-02 DOI: 10.1080/15472450.2023.2209913

Jianli Zhao , Rumeng Zhang , Qiuxia Sun , Jingshi Shi , Futong Zhuo , Qing Li

With the development and acceleration of urbanization, urban metro traffic is gradually growing up to a large network, and the structure of topology between stations becomes more complex, which makes it increasingly difficult to capture the spatial dependency. The vertical and horizontal interlacing of multiple lines makes the stations distributed topologically, and the traditional graph convolution is implemented on the adjacency matrix generated based on a priori knowledge, which cannot reflect the actual spatial dependence between stations. To address these problems, this paper proposes an adaptive graph convolutional network model (Adapt-GCN), which replaces the fixed adjacency matrix obtained from a priori knowledge in the traditional GCN with a trainable adaptive adjacency matrix. This can not only effectively adjust the weights of correlations between adjacent stations, but also adaptively capture the spatial dependencies between non-adjacent stations. This paper uses the Shanghai Metro dataset to verify the effectiveness of the model in improving prediction accuracy and reducing training time.

随着城市化进程的发展和加快，城市地铁交通逐渐成长为一个大型网络，站点间的拓扑结构也变得越来越复杂，使得空间依赖性的捕捉变得越来越困难。多条线的纵横交错使得台站呈拓扑分布，传统的图卷积是基于先验知识生成的邻接矩阵实现的，无法反映台站之间的实际空间依赖性。为了解决这些问题，本文提出了一种自适应图卷积网络模型（Adapt-GCN），该模型将传统GCN中由先验知识获得的固定邻接矩阵替换为可训练的自适应邻接矩阵。这不仅可以有效地调整相邻台站之间的相关权重，而且可以自适应地捕获非相邻台站之间的空间依赖关系。本文使用上海地铁数据集验证了该模型在提高预测精度和减少训练时间方面的有效性。

{"title":"Adaptive graph convolutional network-based short-term passenger flow prediction for metro","authors":"Jianli Zhao , Rumeng Zhang , Qiuxia Sun , Jingshi Shi , Futong Zhuo , Qing Li","doi":"10.1080/15472450.2023.2209913","DOIUrl":"10.1080/15472450.2023.2209913","url":null,"abstract":"<div><div>With the development and acceleration of urbanization, urban metro traffic is gradually growing up to a large network, and the structure of topology between stations becomes more complex, which makes it increasingly difficult to capture the spatial dependency. The vertical and horizontal interlacing of multiple lines makes the stations distributed topologically, and the traditional graph convolution is implemented on the adjacency matrix generated based on a priori knowledge, which cannot reflect the actual spatial dependence between stations. To address these problems, this paper proposes an adaptive graph convolutional network model (Adapt-GCN), which replaces the fixed adjacency matrix obtained from a priori knowledge in the traditional GCN with a trainable adaptive adjacency matrix. This can not only effectively adjust the weights of correlations between adjacent stations, but also adaptively capture the spatial dependencies between non-adjacent stations. This paper uses the Shanghai Metro dataset to verify the effectiveness of the model in improving prediction accuracy and reducing training time.</div></div>","PeriodicalId":54792,"journal":{"name":"Journal of Intelligent Transportation Systems","volume":"28 6","pages":"Pages 806-815"},"PeriodicalIF":2.8,"publicationDate":"2024-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73297920","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0