Coordinated Variable Speed Limit Control for Consecutive Bottlenecks on Freeways Using Multiagent Reinforcement Learning

Most of the current variable speed limit (VSL) strategies are designed to alleviate congestion in relatively short freeway segments with a single bottleneck. However, reality, consecutive bottlenecks can occur simultaneously due merging flow from multiple ramps. In such situations, existing use VSL controllers that operate independently, without considering traffic interactions and differences. this research, we introduced multiagent reinforcement learning-based (MARL-VSL) approach enhance collaboration among controllers. The MARL-VSL employed centralized training decentralized execution structure achieve joint optimal solution for series bottleneck scenarios were simulated modified cell transmission model validate effectiveness proposed strategy. An independent single-agent (ISARL-VSL) feedback-based (feedback-VSL) also applied comparison. Time-varying heterogeneous stemming mainline ramps was loaded into network. results demonstrated achieved superior performance compared baseline methods. reduced total time spent by vehicles 18.01% 17.07% static dynamic scenarios, respectively. control actions more appropriate maintaining smooth its performance. More specifically, significantly improved average driving homogeneity across entire freeway.