2026· Academic Journal of Engineering and Technology Science· Vol 9· 0 citations· 2 references
TL;DR
This study identifies and quantifies traffic flow spatiotemporal characteristics to establish a targeted quantitative indicator system and designs a full-process big data mining framework to complete multi-source data preprocessing and spatiotemporal correlation feature extraction, constructing a traffic flow prediction model that integrates spatiotemporal characteristics.
Abstract
: Intelligent transportation is the core solution to alleviating urban traffic congestion and enhancing traffic governance efficiency. Accurately capturing the spatiotemporal coupling characteristics of traffic flow is key to achieving precise traffic flow prediction. Addressing current issues such as the neglect of spatiotemporal correlations in traffic flow prediction models, low efficiency in intelligent transportation big data mining, insufficient prediction accuracy, and weak generalization capabilities, this study focuses on the core spatiotemporal characteristics of traffic flow—time periodicity, volatility, spatial correlation, and aggregation—while incorporating the "5V" features of multi-source heterogeneous big data in intelligent transportation. It conducts research on big data mining and prediction models. First, it identifies and quantifies traffic flow spatiotemporal characteristics to establish a targeted quantitative indicator system. Second, it designs a full-process big data mining framework to complete multi-source data preprocessing and spatiotemporal correlation feature extraction. Building on this, it improves and optimizes the model by introducing a spatiotemporal attention mechanism based on LSTM and GCN algorithms, constructing a traffic flow prediction model that integrates spatiotemporal characteristics. Finally, the model's performance is validated through case studies. The results demonstrate that the proposed model significantly outperforms traditional models, with a minimum MAPE of 4.87%, a single-prediction time of 0.32 seconds, and strong real-time adaptability and scenario compatibility. This study enriches the theory of spatiotemporal analysis and big data integration in traffic flow, providing scientific decision-making support for intelligent transportation management and road network optimization, with important theoretical and practical application value.
Addressing challenges in smart city traffic management, where industrial traffic generated by sectors such as manufacturing complicates flow prediction and real-time regulation, this study investigates AI-enabled dynamic traffic flow control models and big data governance strategies to improve operational responsiveness and data utilization efficiency. Building upon traffic flow theory, artificial intelligence algorithms, and big data governance frameworks, the study first establishes the conceptual foundations of AI-empowered dynamic traffic flow control and defines key evaluation dimensions, including prediction accuracy, control response speed, traffic efficiency improvement, and data governance efficiency. Subsequently, multi-source traffic data, including road surveillance, vehicle GPS, public transportation operations, and meteorological information, are collected from 10 representative smart cities to construct the multidimensional STBD-2024 dataset covering six traffic scenarios and five congestion levels. A three-dimensional framework of “Multi-source Data Fusion Governance–Intelligent Flow Forecasting–Dynamic Precision Regulation ” is then proposed to enhance data quality through cleaning, fusion, and standardization while enabling short-term traffic prediction and adaptive regulation based on enhanced deep learning models. The proposed framework also provides a scalable computational paradigm for intelligent sensing and wireless data fusion in connected transportation environments, offering methodological support for electromagnetic information acquisition and real-time perception systems in next-generation smart city infrastructures.
Urban traffic flow data constitutes a complex dynamic system, exhibiting both temporal evolution and spatial dependence. In the context of high-density metropolitan areas such as the Greater Bay Area in China and the construction of emerging smart cities, extracting operational patterns from historical traffic observation data and inferring future traffic conditions has become a core issue in urban computing and intelligent transportation. This paper conducts a logic-oriented analysis of urban traffic spatiotemporal data. First, it examines three temporal evolution characteristics: periodicity, trend, and abrupt change. Then, it analyzes how road network topology affects the spatial dependence between road segments. Based on this, the paper explores the coupling mechanism between the temporal and spatial dimensions from the perspective of congestion propagation dynamics and extracts a logical framework for traffic condition prediction. This paper also discusses key challenges such as cross-regional multi-modal analysis, the integration of domain knowledge and data-driven methods, and the shift from open-loop prediction to closed-loop decision support, aiming to provide a structured analytical framework that contributes to understanding the intrinsic mechanisms of urban traffic system operation.
Shuohan Xu· Applied and Computational En...· 0 citations
A prediction model that incorporates multiple attention mechanisms with spatiotemporal graph convolutional networks (HASTGCN) and designs a spatiotemporal map convolution module to collaboratively model the dynamic spatiotemporal connection of traffic flow collaboratively model is used.
Chu-xia Chen· Proceedings of the 3rd Inter...· 0 citations
Traffic jam is currently one of the most critical issues in contemporary cities because of the high rates of population growth, the possession of vehicles, and the insufficient development of the road system. Smart transportation systems (ITS) heavily rely on predicting traffic flow in the future to allow for proactive traffic control, reduce traffic congestion, optimize routes, and ensure increased safety of commuters. With the development of big data analytics, the prediction of traffic flows has been changed greatly since it taps into large amounts of heterogeneous data produced by sensors, GPS, mobile phones, social media, and intelligent vehicles. The paper is a detailed research of how big data analytics have been used to predict traffic flow. It analyses sources of data, analytics, machine learning and deep- learning models and scalable processing frameworks in modern traffic prediction systems. The most popular traditional statistical models, state-of-the-art deep learning methods convolutional neural networks (CNN), recurrent neural networks (RNN), long short-term memory (LSTM), and graph neural networks (GNN) were mentioned through an extensive literature survey. The suggested approach incorporates data preparation, feature detection, model training and performance analysis in a big data ecosystem. It has been experimentally shown that advanced analytics can be effectively implemented to enhance the accuracy of predictions and make them robust. The paper is summarized by a discussion on challenges, limitations as well as future research directions in the prediction of traffic flow using big data.
Unknown authors· International Journal of App...· 0 citations
Accurate traffic flow prediction is the core foundation of Intelligent Transportation Systems (ITS) and urban traffic dynamic optimization. However, existing mainstream prediction models have three critical limitations: poor cross-road transferability, mandatory separate training for individual road sections, and insufficient mining of historical traffic peak information, which jointly restrict prediction accuracy and engineering scalability in practical urban traffic management. To address these challenges, this article proposes the Multi-road Multi-Peak-aware Informer (MMP-Informer), an improved Informer-based long-term time series prediction model with a novel dual-peak-aware attention mechanism. This core module adaptively locates historical maximum and minimum traffic peaks
via
an adaptive window matched to input sequence length, introduces learnable parameters to weight the influence of different peaks on future traffic flow, and realizes single-model end-to-end simultaneous multi-road traffic flow prediction. Meanwhile, the model retains the encoder-decoder architecture of the original Informer, integrates self-attention distillation in the encoder to reduce computational overhead, and adopts a generative decoder paradigm to avoid error accumulation in multi-step prediction. Experiments on the 2018 University of Technology Sydney (UTS) multi-source traffic dataset show that MMP-Informer consistently outperforms state-of-the-art baselines including SCINet, NsTransformer, Transformer, and original Informer across 24-, 48-, and 96-h prediction horizons. For 24-h short-term prediction, it achieves a Mean Absolute Error (MAE) of 25.43 and Mean Absolute Percentage Error (MAPE) of 11.55%; for 96-h long-term prediction, it reaches an MAE of 31.21 and MAPE of 12.93%, where the MAE is 2.51 lower and MAPE is 1.97 percentage points lower than the original Informer, respectively. Ablation tests verify that the dual-peak-aware design significantly enhances the model’s adaptability to both low-traffic and high-traffic scenarios, balancing prediction accuracy and long-term forecasting stability. Supplementary experiments on merged Electricity Transformer Temperature hourly (ETTh1/ETTh2) datasets further confirm its stable and superior performance in non-traffic time series scenarios, with strong cross-scenario generalization ability. The proposed model eliminates the high cost of single-road separate training, breaks through the cross-road transferability bottleneck of traditional models, maintains stable accuracy in both short-term and long-term prediction, and has high practical engineering value for large-scale deployment in urban intelligent traffic management.
As urbanization continues to reshape large cities, traffic congestion remains a persistent challenge for urban transportation systems. Using Beijing as a case study, this paper examines the spatiotemporal evolution of urban traffic congestion from a deep learning perspective based on multi-source data. The results suggest that traffic congestion in Beijing displays a pronounced “dual-peak” pattern associated with daily commuting activities. Compared with the pre-pandemic period, weekday travel demand has generally recovered and in some cases exceeded previous levels, whereas holiday travel remains relatively subdued, accompanied by increasingly concentrated travel behavior. Among the models considered, Long Short-Term Memory (LSTM) performs particularly well in capturing nonlinear variations and temporal dependencies in traffic flow, leading to improved prediction accuracy. Between 2020 and 2025, the congestion index experienced a trajectory of decline, recovery, and subsequent stabilization, a pattern that appears to be associated with the gradual implementation of intelligent traffic management measures. These findings contribute to a better understanding of recent changes in urban traffic dynamics and may offer useful insights for future traffic planning and governance.
Zihan Zhou· Computers and artificial int...· 0 citations
