Journal of Engineering, Project, and Production Management, 2027, 17(1), 2026-50
Predicting Traffic Flow Using Enhanced Long Short-Term Memory
Lecturer, Department of Traffic Management Engineering, Henan Police College, Zhengzhou, 450046, China, E-mail: dong.dhl@outlook.com
Project Management
Received January 21, 2026; revised April 17, 2026; accepted September 5, 2026
Available online October 5, 2026
Abstract: In response to the urgent need for real-time, accurate traffic prediction in urban congestion, this study proposes an enhanced Long Short-Term Memory (LSTM)-based approach that leverages collaboration with cloud platforms. First, based on the LSTM model benchmark, a Convolutional Neural Network (CNN) is used to extract local spatial patterns, a double-layer LSTM is used to capture long-term temporal dependencies, and an attention mechanism is introduced to dynamically weight traffic characteristics over time. Second, the urban transportation network is modeled to better analyze vehicle management. Subsequently, local signal adaptive control is combined with regional signal collaborative control, and a hierarchical collaborative strategy for signal control that considers traffic flow balance across the road network is proposed. The results showed that the proposed method achieved a Root Mean Square Error (RMSE) of 15.8 vehicles per hour (veh/h) within 15 minutes and a Mean Absolute Percentage Error (MAPE) of 6.2%, both of which were significantly better than those of the comparison models. An ablation experiment confirmed that the CNN and attention mechanism reduced prediction errors by 18% and 12%, respectively. The proposed signal control strategy effectively reduced average delay time and reduced the number of overflows by 72%. The research method provides a scalable, high-precision, end-to-end solution for cloud-edge-integrated traffic prediction and collaborative management and control.
Keywords: Cloud platform; long short-term memory (LSTM); convolutional neural network (CNN); attention mechanism; traffic flow; signal control; prediction.
Copyright © Journal of Engineering, Project, and Production Management (EPPM-Journal).
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.
Requests for reprints and permissions at eppm.journal@gmail.com.
Citation: Dong, H. (2027). Predicting Traffic Flow Using Enhanced Long Short-Term Memory. Journal of Engineering, Project, and Production Management, 17(1), 2026-50.
