Reinforcement learning for Traffic Signal Control: Comparison with commercial systems

Abstract

In recent years, Intelligent Transportation Systems are leveraging the power of increased sensory coverage and available computing power to deliver data-intensive solutions achieving higher levels of performance than traditional systems. Within Traffic Signal Control (TSC), this has allowed the emergence of Machine Learning (ML) based systems. Among this group, Reinforcement Learning (RL) approaches have performed particularly well. Given the lack of industry standards in ML for TSC, literature exploring RL often lacks comparison against commercially available systems and straightforward formulations of how the agents operate. Here we attempt to bridge that gap. We propose three different architectures for RL based agents and compare them against currently used commercial systems MOVA, SurTrac and Cyclic controllers and provide pseudo-code for them. The agents use variations of Deep Q-Learning (Double Q Learning, Duelling Architectures and Prioritised Experience Replay) and Actor Critic agents, using states and rewards based on queue length measurements. Their performance is compared in across different map scenarios with variable demand, assessing them in terms of the global delay generated by all vehicles. We find that the RL-based systems can significantly and consistently achieve lower delays when compared with traditional and existing commercial systems.