Optimizing Traffic Signal Control Using Reinforcement Learning

Inefficient traffic light control systems contribute significantly to road congestion, leading to various issues. In this manuscript, we propose a reinforcement learning approach for optimizing traffic light control across diverse scenarios. We developed a customizable crossroads environment, allowing us to simulate multiple settings by altering specific variables. Our method employs the Q-learning algorithm to train the reinforcement learning agent. Finally, in order to validate our approach, we tested it in crossroad environments with moderate and low traffic congestion and obtained satisfactory results.

In *Soft Computing Theories and Applications *