IET Intelligent Transport Systems最新文献

Multispectral pedestrian detection with visible light and infrared images is robust to changes in lighting conditions and therefore is of great importance to numerous applications that require all-day environmental perception. This paper proposes a novel method named FCE-RCNN, which integrates saliency detection as a sub-task and utilizes global information for enhanced feature representation. The approach enhances thermal inputs by incorporating gradients at the raw-data level before feature extraction. Utilizing a dual-stream backbone, a global semantic information extraction module is introduced that combines pooling with horizontal–vertical attention mechanisms, capturing high-quality global semantic information for lower-level feature enrichment and guidance. Additionally, the pedestrian locality enhancement module is designed to enhance spatial locality information of pedestrians through saliency detection. Furthermore, to alleviate the challenges posed by positional shifts between cross-spectral features, deformable convolution is innovatively employed. Experimental results on the KAIST dataset demonstrate that FCE-RCNN significantly improves nighttime detection, achieving a log-average miss rate of 6.92%, outperforming the new method ICAFusion by 0.93%. These results underscore the effectiveness of FCE-RCNN, and the method also maintains competitive inference speed, making it suitable for real-time applications.

The analysis of people's daily activities has played a crucial role in various applications, such as urban geography, activity prediction, and homogeneous population detection. However, limited studies have explored changes in the residents’ activity patterns in a particular region across various periods. To explore the changes, a methodological framework of sequence visualization analysis based on machine learning that extracts the activity patterns across various periods using sequence analysis, visualizes the activity patterns by calculating the frequency of different activities at time points and categorizes them through graphical similarity, and then compares the activity patterns in terms of activity and demographic characteristics is proposed. Empirical testing on the New York Metropolitan data of the National Household Travel Survey (NHTS) is conducted for 2001, 2009, and 2017. The findings reveal significant intra-similarities, inter-differences, and distinct changes in activity patterns across three periods for different social populations in the New York Metropolitan. From the perspective of information analysis, this work is anticipated to enhance the understanding of travel needs for diverse social populations in a particular region, thereby facilitating targeted policy adjustments for the departments concerned.

The study of vessel trajectories (VTs) holds significant benefits for marine route management and resource development. VT segmentation serves as a foundation for extracting vessel motion primitives and enables analysis of vessel manoeuvring habits and behavioural intentions. However, existing methods relying on predefined behaviour patterns face high labelling costs, which hinder accurate pattern recognition. This paper proposes a self-supervised vessel trajectory segmentation method (SS-VTS), which segments VTs based on their inherent spatio-temporal semantics. SS-VTS adaptively divides VTs into cells of optimal size. Then, it extracts split points on different semantic levels from the multi-dimensional feature sequence of the VTs using self-supervised learning. Finally, spatio-temporal distance fusion module is performed on split points to determine change points and obtain VT segments with multiple semantics. Experiments on a real automatic identification system datasets show that SS-VTS achieves state-of-the-art segmentation results compared to seven baseline methods.

Customized bus services typically focus on single-trip requests, which often struggle to accommodate the growing needs for varied multiple trips in urban daily travel. This paper addresses the customized bus routing problem for passengers with multiple trips. A bi-objective mathematical model is established for maximizing the operational profit and minimizing the travel costs by considering the characteristics of the multi-trip requests and time-dependent travel time. Besides, a novel profit objective function is proposed considering the service's completion status and the starting price. Since the proposed mixed integer linear programming model is an NP-hard problem, a non-dominated sorting genetic algorithm II-based method is proposed to handle different sizes of instances. Finally, the instances with multi-trip requests are carried out to test the accuracy of the model and the effectiveness of our method compared with Gurobi and the local search-based multi-objective algorithm approach.

Max-weight (or max-pressure) is a popular traffic signal control algorithm that has been demonstrated to be capable of optimising network-level throughput. It is based on queue size measurements in the roads approaching an intersection. However, the inability of typical sensors to accurately measure the queue size results in noisy queue measurements, which may affect the controller's performance. A possible solution is to utilise the noisy max-weight algorithm to achieve a throughput optimal solution; however, its application may lead to decreased controller performance. This article investigates two variants of max-weight controllers, namely, acyclic and cyclic max-weight (with and without noisy queue information) in simulated scenarios, by examining their impact on the throughput and environment. A detailed study of multiple pollutants, fuel consumption, and traffic conditions, which are proxied by a total social cost function, is presented to show the pros and cons of each controller. Simulation experiments, conducted for the Kamppi area in central Helsinki, Finland, show that the acyclic max-weight controller outperforms a fixed time controller, particularly in avoiding congestion and reducing emissions in the network, while the cyclic max-weight controller gives the best performance to accommodate maximum vehicles flowing in the network. The complementary positive characteristics motivated the authors to propose a new controller, herein called the hybrid max-weight, which integrates the characteristics of both acyclic and cyclic max-weight algorithms for providing better traffic load and performance through switching.

To promote urban sustainability, many cities are adopting bicycle-friendly policies, leveraging GPS trajectories as a vital data source. However, the inherent errors in GPS data necessitate a critical preprocessing step known as map-matching. Due to GPS device malfunction, road network ambiguity for cyclists, and inaccuracies in publicly accessible streetmaps, existing map-matching methods face challenges in accurately selecting the best-mapped route. In urban settings, these challenges are exacerbated by high buildings, which tend to attenuate GPS accuracy, and by the increased complexity of the road network. To resolve this issue, this work introduces a map-matching method tailored for cycling travel data in urban areas. The approach introduces two main innovations: a reliable classification of road availability for cyclists, with a particular focus on the main road network, and an extended multi-objective map-matching scoring system. This system integrates penalty, geometric, topology, and temporal scores to optimize the selection of mapped road segments, collectively forming a complete route. Rotterdam, the second-largest city in the Netherlands, is selected as the case study city, and real-world data is used for method implementation and evaluation. Hundred trajectories were manually labelled to assess the model performance and its sensitivity to parameter settings, GPS sampling interval, and travel time. The method is able to unveil variations in cyclist travel behavior, providing municipalities with insights to optimize cycling infrastructure and improve traffic management, such as by identifying high-traffic areas for targeted infrastructure upgrades and optimizing traffic light settings based on cyclist waiting times.

The intelligent expressway exemplifies a prominent application of intelligent transportation systems. Roadside units (RSUs), strategically deployed alongside roadways, serve as pivotal infrastructure in facilitating interactions within intelligent expressways. A well-planned RSU deployment strategy is crucial for enhancing service quality, it necessitates balancing performance improvements with significant financial costs due to the limited transmission range and high deployment expenses of RSUs. To tackle these challenges, an adaptive approach for RSU deployment is proposed, which takes into account economic feasibility, service requirements, and dynamic traffic demands. A traffic adaptability-based RSU deployment (TARD) model, which integrates factors such as deployment cost, the effectiveness of information coverage, road network topology, and traffic flow characteristics have been devised. The TARD aims to minimize deployment expenses while maximizing the benefits of information coverage and alignment with road traffic demands. The Non-dominated Sorting Genetic Algorithm II (NSGA-II) is employed to solve this optimization model. To validate its efficacy, simulations are conducted on the G2 expressway in Shandong Province, China, demonstrating the superior performance of the TARD compared to three other deployment strategies. Ablation experiments further underscore the critical role of tunnel deployments and comprehensive coverage along long sections in bolstering network connectivity and elevating service quality.

Drawing inspiration from the state-of-the-art object detection framework YOLOv8, a new model termed adverse weather net (ADWNet) is proposed. To enhance the model's feature extraction capabilities, the efficient multi-scale attention (EMA) module has been integrated into the backbone. To address the problem of information loss in fused features, Neck has been replaced with RepGDNeck. Simultaneously, to expedite the model's convergence, the bounding box's loss function has been optimized to SIoU loss. To elucidate the advantages of ADWNet in the context of adverse weather conditions, ablation studies and comparative experiments were conducted. The results indicate that although the model's parameter count increased by 18.4%, the accuracy for detecting rain, snow, and fog in adverse weather conditions improved by 22%, while the FLOPs (floating point operations) decreased by 5%. The results of the comparison experiments conducted on the WEDGE dataset show that ADWNet outperforms other object detection models in adverse weather in terms of accuracy, model parameters and FLOPs. To validate ADWNet's real-world efficacy, data was extracted from a car recorder under adverse conditions on highways, visual inference was conducted, and its accuracy was demonstrated in interpreting real-world scenarios. The config files are available at https://github.com/Xinyun-Feng/ADWNet.

The UK rail network is subject to costly disruption due to the operational effects of adhesion variation between the wheel and rail. Causes of this are often environmental introduction of contaminants that require a wide-scale approach to risk mitigation such as defensive driving or rail-head maintenance. It remains an open problem to monitor the real-time status of the network to optimise resources and approaches in response to adhesion problems. This article presents an on-vehicle monitoring method designed to estimate the coefficient of friction by processing data from on-board sensors of typical rail passenger vehicles. This approach uses a multi-body physics analysis of a target vehicle to create estimators for both creep force and creep, allowing a curve fitting approach to estimate the coefficient for friction from the creep curves.

Cycling is increasingly promoted worldwide, but many urban areas lack satisfactory cycling environments. Assessing these environments is crucial, but existing methods face data challenges for large urban networks. This study proposes a data-driven framework using dockless shared bicycle data to efficiently evaluate large-scale cycling environments. First, critical cycling behaviour features that reflect cyclists’ perceptions are identified applying the fuzzy C-means and random forest model. Then, a distribution-oriented evaluation method is developed, ensuring the incorporation of cyclist heterogeneity and quantifying the quality differences among road segments by combining statistical analysis with a hierarchical clustering model. The evaluation framework is applied to Yangpu District, Shanghai, using Mobike data covering 114.9 km of cycling roads. Results show that indicators related to speed magnitude and fluctuation are critical, and an experimental study validates the effectiveness of the data-driven feature extraction method. A minimum trajectory sample size of 260 is required to account for cyclist heterogeneity for one road segment to be evaluated. Further analysis of lower-performing segments identifies vehicle-bicycle separation, on-street parking, and traffic volume as key influencing factors. The rationality of these findings further supports the reliability of the evaluation framework.