Multi-Objective Adaptive Traffic Signal Control Using Fuzzy Control and Q-Learning

A multi-objective adaptive traffic signal control algorithm using fuzzy control and Q-learning was proposed to improve the efficiency, traffic safety, and operational stability of signalized intersections. In this algorithm, the signal cycle length was derived by fuzzy control, then, to minimize del...

Full description

Saved in:
Bibliographic Details
Published in:2024 12th International Conference on Traffic and Logistic Engineering (ICTLE) pp. 57 - 62
Main Authors: Ding, Naikan, Ma, Zufan, Lu, Zhaoyou, Wan, Chengpeng
Format: Conference Proceeding
Language:English
Published: IEEE 23.08.2024
Subjects:
adaptive traffic signal control
Computer crashes
Delays
Fuzzy control
Heuristic algorithms
Information entropy
Logistics
multi-objective optimization
Q-learning
Safety
Simulation
Stability analysis
traffic simulation
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:A multi-objective adaptive traffic signal control algorithm using fuzzy control and Q-learning was proposed to improve the efficiency, traffic safety, and operational stability of signalized intersections. In this algorithm, the signal cycle length was derived by fuzzy control, then, to minimize delay and conflicts, the green split of each phase was dynamically adjusted through Q-learning. A joint simulation of Python and VISSIM was adopted for traffic operational simulation and evaluation. The simulation results show that the proposed algorithm jointing fuzzy control and Q-learning, and compared with traffic actuated control and fixed timing, the delay, queue length and traffic conflict of the intersection are significantly and comprehensively optimized. In addition, the algorithm reduced the platoon crash risk at the intersection, improving the overall operational stability.
DOI:10.1109/ICTLE62418.2024.10703932