日本フィジカルAI新聞

世界のフィジカルAIを、日本語で。

週刊ニュースレター購読
マルチエージェント強化学習arXiv:2609.07618

予測シールディングによる分散型安全マルチエージェント強化学習

Decentralized Safe Multi-Agent Reinforcement Learning via Predictive Shielding

シェア:XThreadsFacebookLINEはてブBluesky

複数ロボットが独立にタスクを実行する環境で、訓練時と異なる状態に直面しても安全に適応できるよう、予測シールディングとモデルベースの有限地平Q学習を組み合わせた分散型フレームワークを提案した。対称シナリオでのライブロックを緩和する通信不要のプロトコルも導入している。

著者: Yacine El Yamani, Hanna Krasowski, Elena Vanneaux

分類: eess.SY, cs.AI, cs.MA, cs.RO

原文アブストラクト

Environments are increasingly populated by multiple robots performing independent tasks with limited prior knowledge of each other. Deploying such multi-agent systems presents significant challenges. Specifically, shifts in deployment states compared to training data can lead to poor policy performance and compromised safety. While safety shields exist to mitigate these risks, they are typically reactive, which degrades performance near unseen obstacles,and centralized, limiting their scalability. To address this, we propose a decentralized framework that integrates predictive shielding with model-based finite horizon Q-learning. This approach allows agents to safely adapt their pre-trained policies during deployment. Furthermore, to mitigate livelocks in symmetric scenarios, we introduce a communication- free protocol for conflict resolution

関連論文