An adaptive graph reinforcement learning method for scalable multi‐train cooperative control

Multi‐Train Optimal Control (MTOC) addresses the cooperative control problem of multi‐trains running on railway tracks through centralized or distributed controllers. However, two critical challenges emerge in solving MTOC problems: (1) the dynamic system dimensionality caused by time‐varying train numbers during station arrivals and departures and (2) the strong inter‐train command correlations in dense traffic scenarios. These complexities lead to computational challenges when scaling to extended railway networks with growing train populations, rendering conventional rule‐based methods ineffective. To address these challenges, we propose Graph Attention Soft Actor‐Critic (GASAC), a novel graph reinforcement learning algorithm integrating two core components: (1) A graph attention network (GAT) for efficient information aggregation from high‐dimensional train observations, and (2) A Soft Actor‐Critic (SAC) architecture serving as the centralized decision‐maker. The GAT module performs dimensionality reduction through feature attention mechanisms, effectively supporting the SAC module in deriving optimal control policies. Comparative evaluations against multi‐agent deep reinforcement learning baselines demonstrate that GASAC successfully synthesizes distributed train information to generate control commands, ensuring collision‐free and on‐time operations. Further sensitivity analysis shows the adaptability of the algorithm to different parameters.

• Publication year

  2025

• Venue

  Comput. Aided Civ. Infrastructure Eng.

• Publication date

  2025-11-10

• Fields of study

  Computer Science, Engineering

• Identifiers
  DOI 10.1111/mice.70132
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• MV-STGHAT: Multi-View Spatial-Temporal Graph Hybrid Attention Network for Decision-Making of Connected and Autonomous Vehicles

  2025cited by this paper
• PyG 2.0: Scalable Learning on Real World Graphs

  2025cited by this paper
• A reinforcement learning approach to solving very-short term train rescheduling problem for a single-track rail corridor

  2024cited by this paper
• Graph-based multi-agent reinforcement learning for collaborative search and tracking of multiple UAVs

  2024cited by this paper
• Deep Deterministic Policy Gradient Virtual Coupling control for the coordination and manoeuvring of heterogeneous uncertain nonlinear High-Speed Trains

  2024cited by this paper
• Deterministic reinforcement learning for optimized formation control of virtually-coupled trains via performance index monitor

  2023cited by this paper
• Railway Virtual Coupling: A Survey of Emerging Control Techniques

  2023influential reference
• Cooperative train control during the power supply shortage in metro system: A multi-agent reinforcement learning approach

  2023cited by this paper
• DDPG-Based Energy-Efficient Train Speed Trajectory Optimization Under Virtual Coupling

  2023cited by this paper
• Research on the stability of virtual coupling train formation based on PID controller under the influence of time delay

  2023cited by this paper
• A Hybrid Online Off-Policy Reinforcement Learning Agent Framework Supported by Transformers

  2023cited by this paper
• Reaching the limit in autonomous racing: Optimal control versus reinforcement learning

  2023cited by this paper
• Optimal Motion of a Rolling Stock Fleet under Traction Power System Constraints

  2022cited by this paper
• Research on headway control of guided transport system based on intersections conditions evaluation

  2022cited by this paper
• New adaptive robust H∞ control of smart structures using synchrosqueezed wavelet transform and recursive least-squares algorithm

  2022cited by this paper
• STGC-GNNs: A GNN-based traffic prediction framework with a spatial-temporal Granger causality graph

  2022cited by this paper
• New adaptive robust <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" display="inline" id="d1e4710" altimg="si5.svg"><mml:msub><mml:mrow><mml:mi>H</mml:mi></mml:mrow><mml:mrow><mml:mi>∞</mml:mi></mml:mrow></mml:msub></mml:math> control of smart structures using synchrosqueezed wavelet transfo

  2022cited by this paper
• STGNN-TTE: Travel time estimation via spatial-temporal graph neural network

  2022cited by this paper
• Energy-efficient operation by cooperative control among trains: A multi-agent reinforcement learning approach

  2021cited by this paper
• STGSN - A Spatial-Temporal Graph Neural Network framework for time-evolving social networks

  2021cited by this paper
• The Surprising Effectiveness of PPO in Cooperative Multi-Agent Games

  2021cited by this paper
• A Virtually Coupled Metro Train Platoon Control Approach Based on Model Predictive Control

  2021cited by this paper
• An analytical optimal control approach for virtually coupled high-speed trains with local and string stability

  2021cited by this paper
• Graph neural network and reinforcement learning for multi‐agent cooperative control of connected autonomous vehicles

  2021cited by this paper
• A Model Predictive Control Based Inter-Station Driving Strategy for Virtual Coupling Trains in Railway System

  2021cited by this paper
• Message-Aware Graph Attention Networks for Large-Scale Multi-Robot Path Planning

  2020cited by this paper
• A Model Predictive Control Approach for Virtual Coupling in Railways

  2019cited by this paper
• Graph Neural Networks for Decentralized Multi-Robot Path Planning

  2019cited by this paper
• A Comprehensive Survey on Graph Neural Networks

  2019cited by this paper
• Learning Steady-States of Iterative Algorithms over Graphs

  2018cited by this paper
• GeniePath: Graph Neural Networks with Adaptive Receptive Paths

  2018cited by this paper
• GaAN: Gated Attention Networks for Learning on Large and Spatiotemporal Graphs

  2018cited by this paper
• Soft Actor-Critic Algorithms and Applications

  2018cited by this paper
• Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor

  2018cited by this paper
• Graph Attention Networks

  2017cited by this paper
• Joint optimization of high-speed train timetables and speed profiles: A unified modeling approach using space-time-speed grid networks

  2017cited by this paper
• Multi-agent replicator controller for sustainable vibration control of smart structures

  2017cited by this paper
• Deep Neural Networks for Learning Graph Representations

  2016cited by this paper
• New discrete-time robust H2/H∞ algorithm for vibration control of smart structures using linear matrix inequalities

  2016cited by this paper
• Structural Deep Network Embedding

  2016cited by this paper
• Gated Graph Sequence Neural Networks

  2015cited by this paper
• T test as a parametric statistic

  2015cited by this paper
• Spectral Networks and Locally Connected Networks on Graphs

  2013cited by this paper
• Neural Network for Graphs: A Contextual Constructive Approach

  2009cited by this paper
• Inequalities

  2006cited by this paper
• A new model for learning in graph domains

  2005cited by this paper
• Supervised neural networks for the classification of structures

  1997cited by this paper

• No citing papers are available for this paper.