| Predictive Modeling of Vehicle Speed in Urban Traffic Using LiDAR-Measured Following Distance |
| Wookeun Choi1, Seongsu Kim1, Jingyu Yang1, Junghwan Kim2 |
1Department of Energy Systems Engineering, Chung-Ang University, Seoul, 06974, Korea
2School of Energy Systems Engineering, Chung-Ang University, Seoul, Korea |
|
|
|
Received: June 9, 2025; Revised: August 8, 2025 Accepted: September 9, 2025. Published online: November 5, 2025. |
|
|
|
| ABSTRACT |
|
This study investigates the relationship between vehicle speed and following distance using real-world driving data collected via global positioning system (GPS) and light detection and ranging (LiDAR) sensors in urban traffic. Results show that following distance (FD) increases with vehicle speed, reflecting driver comfort and perceived safety. Speed enforcement cameras and road curvature were found to significantly affect driver behavior, often leading to congestion. A vehicle speed prediction model was developed using a Fine Tree regression algorithm, incorporating features such as following distance, road curvature, speed limit, and driver’s comfort speed (DCS). The model achieved strong performance (R² = 0.80, RMSE = 11.62 km/h) and was validated on unseen data. These findings offer practical insights for intelligent traffic systems and lay the groundwork for future modeling of following distance under varying traffic conditions. |
| Key Words:
Vehicle speed · Prediction model · Driving behaviour · Traffic conditions · LiDAR |
|
Corresponding Author.
Junghwan Kim , Email. 
Preview 
Full text via DOI 
Download Citation
Print
