Article
Green & Sustainable Science & Technology
Uchenna Nnabuihe Uhegbu, Miles R. Tight
Summary: The study investigated road user attitudes and behaviors in Abuja, Nigeria, finding that the majority of road users showed high non-compliance with seatbelt usage and also engaged in drink driving, using mobile phones while driving, and not using child restraints. Recommendations include stricter enforcement of road safety laws and providing road safety agents with necessary equipment.
Article
Computer Science, Information Systems
Yu Sang, Zhenping Xie, Wei Chen, Lei Zhao
Summary: Linking trajectories to users who generate them has been a popular research topic due to the large-scale trajectory data obtained by GPS-enabled devices and the widespread applications served by the study. In this work, we propose a novel semi-supervised model, TULRN, based on Graph Neural Network and BiLSTM, which extends the trajectory user linking problem onto road networks and utilizes both the structure of road networks and the sequential characteristics of trajectories. Experimental results on a real-world dataset demonstrate that our proposed TULRN model outperforms state-of-the-art methods.
WORLD WIDE WEB-INTERNET AND WEB INFORMATION SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Guido Novati, Hugues Lascombes de Laroussilhe, Petros Koumoutsakos
Summary: Researchers have used multi-agent reinforcement learning to improve the discovery of turbulence models, with promising results. This approach can estimate unresolved subgrid-scale physics and generalize well across different grid sizes and flow conditions.
NATURE MACHINE INTELLIGENCE
(2021)
Article
Computer Science, Artificial Intelligence
Lingxiao Li, Muhammad Aamir Cheema, Hua Lu, Mohammed Eunus Ali, Adel N. Toosi
Summary: This study compares the quality of alternative routes generated by four popular approaches on road networks in Melbourne, Dhaka, and Copenhagen. The results show that there is no significant difference in the average ratings of these approaches.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2022)
Article
Computer Science, Interdisciplinary Applications
Pin Zhang, Zhen-Yu Yin, Brian Sheil
Summary: There is great potential for machine learning to improve constitutive modelling of geomaterials. However, a lack of interpretability and heavy reliance on big data has been a common criticism. This study proposes an interpretable data-driven approach for geotechnical modelling, incorporating prior knowledge and uncertainty. By adopting a multi-fidelity modelling framework, the impact of small datasets can be maximized. The results show that data-driven modelling with physical constraints performs robustly, even for extrapolation beyond the original dataset.
COMPUTERS AND GEOTECHNICS
(2023)
Article
Environmental Studies
Pouya Zangeneh, Brenda McCabe
Summary: This paper proposes an expandable Object-Oriented Bayesian Network (OOBN) model to incorporate the effects and dependencies of project risk factors and outcome variables related to cost. The methodology helps overcome various biases within project estimation processes by combining both singular-evidence of the project at hand (i.e., inside view) and distributional-evidence of the peer reference class (i.e., outside view).
Article
Environmental Sciences
An Wang, Junshi Xu, Ran Tu, Mingqian Zhang, Matthew Adams, Marianne Hatzopoulou
Summary: The study proposed an integrated traffic emission-dispersion modelling chain that incorporates major sources of uncertainty, generating PM2.5 probability distributions and validating modelled levels. A policy scenario revealed that upgrading 55% of current trucks could lead to a 30% reduction in near-road PM2.5 concentrations.
ENVIRONMENTAL POLLUTION
(2021)
Article
Engineering, Electrical & Electronic
Qinghan Sun, Qun Chen
Summary: This paper proposes a fully decentralized optimization method for multi-agent distribution networks. By introducing the flexibility boundaries and flexibility transition model, the adjustability of nodes and their neighborhood relationship are considered. A robust interval power flow model is established to consider the impact of uncertainties on nodal voltage. The problem is solved using ADMM without global information, resulting in a robust optimization solution that meets system requirements.
IEEE TRANSACTIONS ON SMART GRID
(2022)
Article
Computer Science, Artificial Intelligence
Wenyi Tang, Bei Hui, Ling Tian, Guangchun Luo, Zaobo He, Zhipeng Cai
Summary: User representation learning is a critical task in social network analysis, with the proposed adversarial fusion framework utilizing multi-view information for robust and interpretable user representations. The framework includes a generator and discriminator, with the generator using a variational autoencoder to capture and disentangle latent factors behind user intentions. Extensive experiments on both synthetic and real-world datasets demonstrate the superiority of the proposed model.
INFORMATION FUSION
(2021)
Article
Engineering, Environmental
Shuai Han, Heng Li, Mingchao Li, Jiawen Zhang, Runhao Guo, Jie Ma, Wenchao Zhao
Summary: This article introduces a novel solution for uncertainty analysis and three-dimensional modeling using deep learning methods. It presents two methods, namely spatial uncertainty perception (SUP) and graph representation (GRep), and applies them to a practical engineering project.
BULLETIN OF ENGINEERING GEOLOGY AND THE ENVIRONMENT
(2022)
Article
Computer Science, Interdisciplinary Applications
Sam O'Neill, Ovidiu Bagdasar, Stuart Berry, Nicolae Popovici, Ramachandran Raja
Summary: This paper presents a method of considering multiple objectives simultaneously in selfish routing of network flow. By manipulating free parameters such as speed limits, the behavior of road users is coerced to reconcile conflicts between multiple objectives. The results show that small parameter adjustments can lead to solutions that Pareto dominate other solutions.
MATHEMATICS AND COMPUTERS IN SIMULATION
(2022)
Article
Engineering, Electrical & Electronic
Huaxin Pei, Yi Zhang, Qinghua Tao, Shuo Feng, Li Li
Summary: In this paper, a distributed strategy is proposed to decompose the problem of cooperative driving in multi-intersection road networks into small-scale sub-problems and ensure appropriate coordination between adjacent areas through specially designed information exchange. Simulation results demonstrate the efficiency-complexity balanced advantage of the proposed strategy under various traffic demand settings.
IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY
(2021)
Article
Environmental Studies
Adarsh Yadav, Jyoti Mandhani, Manoranjan Parida, Brind Kumar
Summary: This study develops a traffic noise prediction model using an integrated Bayesian networks and Partial least squares structural equation modeling approach and identifies the direct and indirect effects of five influential factors on traffic noise. Traffic flow attributes and honking are the most crucial factors in degrading the acoustical climate at intersections.
TRANSPORTATION RESEARCH PART D-TRANSPORT AND ENVIRONMENT
(2022)
Article
Computer Science, Information Systems
Haitao Yuan, Sai Wang, Zhifeng Bao, Shangguang Wang
Summary: In this paper, we propose an end-to-end neural network model to improve the accuracy of road extraction by addressing three significant issues: the lack of complementarity among multiple data sources, rough edges of extracted roads, and false positives caused by confusing pixels. Our model leverages encoding networks, attention mechanism, auxiliary tasks, and pixel-aware contrastive-learning module to achieve better results. Additionally, we introduce a model-agnostic transfer learning method to enhance the learning effectiveness of our model.
PROCEEDINGS OF THE VLDB ENDOWMENT
(2023)
Article
Computer Science, Artificial Intelligence
Ming-Wen Shao, Le Li, De-Yu Meng, Wang-Meng Zuo
Summary: In this study, a new method for raindrop removal is proposed, utilizing a soft mask and multi-scale fusion representation to deal with raindrops of different blurring degrees and resolutions, improving the removal effectiveness.
IEEE TRANSACTIONS ON IMAGE PROCESSING
(2021)
