Related Experiment Video
Updated: Apr 20, 2026
06:37
Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention
Published on: December 15, 2023
5.8K
A red-light running prevention system based on artificial neural network and vehicle trajectory data
Pengfei Li1, Yan Li2, Xiucheng Guo3
1Transportation School, Southeast University, 2 Sipailou, Nanjing 210096, China ; School of Computing, Informatics and Decision Systems Engineering, Arizona State University, Tempe, AZ 85281, USA.
Computational Intelligence and Neuroscience
|December 2, 2014
Summary
This study introduces an artificial neural network (ANN) system to predict and prevent red-light running by analyzing complex driver behaviors at intersections. The novel system achieves over 80% accuracy, enhancing traffic safety.
Area of Science:
- Traffic Engineering
- Artificial Intelligence
- Behavioral Science
Background:
- Red-light running is a significant safety concern at intersections.
- Existing models struggle to explain complex driver behaviors during yellow and all-red intervals.
- The dilemma zone and vehicle kinematics alone do not fully account for red-light running frequency.
Purpose of the Study:
- To develop an innovative red-light running prevention system.
- To approximate complex driver behaviors using artificial neural networks (ANNs).
- To identify potential red-light runners based on ANN predictions and vehicle trajectory.
Main Methods:
- Utilized artificial neural networks (ANNs) to model driver behavior during yellow and all-red clearance phases.
- Integrated ANNs with vehicle trajectory analysis for runner identification.
- Developed and tested a prototype red-light running prevention system.
Main Results:
- The ANN model accurately approximated complex driver behaviors.
- The system achieved a prediction accuracy rate exceeding 80%.
- ANN training time was found to be acceptable for practical application.
Conclusions:
- The developed ANN-based system effectively identifies potential red-light runners.
- The system offers a novel approach to red-light running prevention.
- The prototype system is designed for seamless integration with existing traffic signal infrastructure.