CycLight: learning traffic signal cooperation with a cycle-level strategy
This addresses traffic management efficiency for urban planners and commuters, representing an incremental improvement with a novel method for a known bottleneck.
The study tackled network-level adaptive traffic signal control by introducing CycLight, a cycle-level deep reinforcement learning approach that optimizes cycle length and splits simultaneously, resulting in superior performance and robustness against delays compared to state-of-the-art methods in simulations.
This study introduces CycLight, a novel cycle-level deep reinforcement learning (RL) approach for network-level adaptive traffic signal control (NATSC) systems. Unlike most traditional RL-based traffic controllers that focus on step-by-step decision making, CycLight adopts a cycle-level strategy, optimizing cycle length and splits simultaneously using Parameterized Deep Q-Networks (PDQN) algorithm. This cycle-level approach effectively reduces the computational burden associated with frequent data communication, meanwhile enhancing the practicality and safety of real-world applications. A decentralized framework is formulated for multi-agent cooperation, while attention mechanism is integrated to accurately assess the impact of the surroundings on the current intersection. CycLight is tested in a large synthetic traffic grid using the microscopic traffic simulation tool, SUMO. Experimental results not only demonstrate the superiority of CycLight over other state-of-the-art approaches but also showcase its robustness against information transmission delays.