Graph Neural Networks with Model-based Reinforcement Learning for Multi-agent Systems

TL;DR

This paper introduces a novel approach combining Graph Neural Networks with Model-based Reinforcement Learning to improve decision-making in multi-agent systems, demonstrated through tasks like billiard-avoidance and autonomous driving.

Contribution

The paper proposes a new GNN-based MBRL model specifically designed for multi-agent systems, integrating state prediction and planning optimization.

Findings

Effective state and trajectory prediction for multiple agents

Successful task completion in billiard-avoidance and autonomous driving scenarios

Demonstrated improvement over traditional methods in multi-agent planning

Abstract

Multi-agent systems (MAS) constitute a significant role in exploring machine intelligence and advanced applications. In order to deeply investigate complicated interactions within MAS scenarios, we originally propose "GNN for MBRL" model, which utilizes a state-spaced Graph Neural Networks with Model-based Reinforcement Learning to address specific MAS missions (e.g., Billiard-Avoidance, Autonomous Driving Cars). In detail, we firstly used GNN model to predict future states and trajectories of multiple agents, then applied the Cross-Entropy Method (CEM) optimized Model Predictive Control to assist the ego-agent planning actions and successfully accomplish certain MAS tasks.