In urban environments, traffic congestion remains a pervasive challenge, significantly impacting both the economy and quality of life. This issue is addressed in this paper by leveraging Reinforcement Learning (RL) techniques to optimize traffic signal timings at intersections. The objective is to minimize congestion and enhance overall traffic flow. This involves the creation of a simulation environment modelling real-world traffic scenarios. Within this framework, an RL agent is trained to learn optimal signal timings by iteratively interacting with simulated traffic conditions. This proposed work evaluates the trained model across various scenarios, assessing its adaptability to changing traffic conditions and its effectiveness in minimizing congestion. Results from simulations will demonstrate the potential of RL-based traffic signal optimization in significantly improving traffic flow, reducing congestion-related delays, and enhancing the overall efficiency of urban transportation systems. The findings contribute to the ongoing discourse on intelligent traffic management and underscore the applicability of RL techniques in addressing complex real- world challenges.
Smart Traffic Control: Leveraging DQN for Improved Urban Mobility
2024-08-02
576546 byte
Conference paper
Electronic Resource
English