Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

DVS-RG: Differential Variable Speed Limits Control using Deep Reinforcement Learning with Graph State Representation

Jingwen Yang, Ping Wang, Fatemeh Golpayegani, Shen Wang

TL;DR

This work tackles variable speed limit control by introducing a topology-aware differential VSL framework (DVS-RG) that uses a graph-state representation of the road network. It combines lane-level traffic features with graph structure and optimizes continuous, per-lane speed limits via PPO in a graph-based DRL setting. A novel reward normalization and graph-message-passing state representation enable stable training and effective policy learning, achieving substantial improvements in efficiency (e.g., up to 68.44% reduction in average waiting time) and safety (around 15.93% reduction in potential collisions) on SUMO-based bottleneck scenarios. The approach demonstrates the value of incorporating network topology into DRL for DVSL and offers an open-source implementation for broader adoption.

Abstract

Variable speed limit (VSL) control is an established yet challenging problem to improve freeway traffic mobility and alleviate bottlenecks by customizing speed limits at proper locations based on traffic conditions. Recent advances in deep reinforcement learning (DRL) have shown promising results in solving VSL control problems by interacting with sophisticated environments. However, the modeling of these methods ignores the inherent graph structure of the traffic state which can be a key factor for more efficient VSL control. Graph structure can not only capture the static spatial feature but also the dynamic temporal features of traffic. Therefore, we propose the DVS-RG: DRL-based differential variable speed limit controller with graph state representation. DVS-RG provides distinct speed limits per lane in different locations dynamically. The road network topology and traffic information(e.g., occupancy, speed) are integrated as the state space of DVS-RG so that the spatial features can be learned. The normalization reward which combines efficiency and safety is used to train the VSL controller to avoid excessive inefficiencies or low safety. The results obtained from the simulation study on SUMO show that DRL-RG achieves higher traffic efficiency (the average waiting time reduced to 68.44\%) and improves the safety measures (the number of potential collision reduced by 15.93\% ) compared to state-of-the-art DRL methods.

DVS-RG: Differential Variable Speed Limits Control using Deep Reinforcement Learning with Graph State Representation

TL;DR

This work tackles variable speed limit control by introducing a topology-aware differential VSL framework (DVS-RG) that uses a graph-state representation of the road network. It combines lane-level traffic features with graph structure and optimizes continuous, per-lane speed limits via PPO in a graph-based DRL setting. A novel reward normalization and graph-message-passing state representation enable stable training and effective policy learning, achieving substantial improvements in efficiency (e.g., up to 68.44% reduction in average waiting time) and safety (around 15.93% reduction in potential collisions) on SUMO-based bottleneck scenarios. The approach demonstrates the value of incorporating network topology into DRL for DVSL and offers an open-source implementation for broader adoption.

Abstract

Variable speed limit (VSL) control is an established yet challenging problem to improve freeway traffic mobility and alleviate bottlenecks by customizing speed limits at proper locations based on traffic conditions. Recent advances in deep reinforcement learning (DRL) have shown promising results in solving VSL control problems by interacting with sophisticated environments. However, the modeling of these methods ignores the inherent graph structure of the traffic state which can be a key factor for more efficient VSL control. Graph structure can not only capture the static spatial feature but also the dynamic temporal features of traffic. Therefore, we propose the DVS-RG: DRL-based differential variable speed limit controller with graph state representation. DVS-RG provides distinct speed limits per lane in different locations dynamically. The road network topology and traffic information(e.g., occupancy, speed) are integrated as the state space of DVS-RG so that the spatial features can be learned. The normalization reward which combines efficiency and safety is used to train the VSL controller to avoid excessive inefficiencies or low safety. The results obtained from the simulation study on SUMO show that DRL-RG achieves higher traffic efficiency (the average waiting time reduced to 68.44\%) and improves the safety measures (the number of potential collision reduced by 15.93\% ) compared to state-of-the-art DRL methods.
Paper Structure (20 sections, 15 equations, 6 figures, 2 tables, 1 algorithm)

This paper contains 20 sections, 15 equations, 6 figures, 2 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Background
  3. Related work
  4. Scope and Contributions
  5. Paper Organization
  6. System Overview
  7. System Design Goals
  8. System Architecture
  9. Action
  10. Reward
  11. State
  12. Deep Reinforcement Learning Algorithm
  13. Evaluation Methodology
  14. Testing Scenario
  15. Evaluation Metrics
  16. ...and 5 more sections

Figures (6)

  • Figure 1: (a) The traffic bottleneck performance under NO-VSL; (b) The traffic bottleneck performance under VSL; (c) The traffic bottleneck performance under DVSL.
  • Figure 2: The actor-critic architecture for DVSL with graph state representation
  • Figure 3: Area divide and detectors deploy of state space
  • Figure 4: The traffic traffic performance of MA.
  • Figure 5: Training process of DVS-PPO, DVS-TD3, DVS-DDPG and DVS-RG methods.
  • ...and 1 more figures