基于扩散的多智能体STL规划方法
原标题：Generalizable Multi-Agent Planning from Signal Temporal Logic Specifications via Diffusion
AI 摘要
本文提出了一种基于扩散模型的多智能体规划方法 Diff-MA，用于满足信号时序逻辑（STL）规范。该方法通过可微分的 STL 近似在去噪过程中集成梯度，实现了对新公式的泛化，同时保持了与现有学习方法相同的可扩展性，并支持异构规范和团队级规范。实验表明，该方法生成的计划具有多样性，显著减少了智能体间的碰撞等安全违规。
正文节选
Generalizable Multi-Agent Planning from Signal Temporal Logic Specifications via Diffusion Abstract Multi‑agent systems in the real-world (e.g., drone swarms, autonomous cars, warehouse robots) must satisfy rich, temporal tasks while avoiding collisions. Signal Temporal Logic (STL) elegantly encodes such objectives, but current STL planning methods face critical limitations. State-of-the-art optimization-based approaches can handle arbitrary STL specifications but struggle with scalability, beco