构建可复现的零样本迁移评估平台，验证自动驾驶多智能体强化学习在真实场景中的表现。
Zero-Shot MARL Benchmark in the Cyber-Physical Mobility Lab
- 基于物理-数字混合测试平台，实现仿真到现实的零样本迁移评估。
- 发现性能下降源于控制栈差异与环境真实性提升双重影响。
- 开源方案适合研究仿真到现实迁移问题的科研人员使用。
我们提出一个可复现的基准测试平台，用于评估多智能体强化学习（MARL）策略在联网自动驾驶车辆（CAVs）中的仿真到现实迁移能力。该平台基于网络物理移动实验室（CPM Lab）[1]，整合了仿真、高保真数字孪生和物理测试平台，支持对MARL运动规划策略进行结构化的零样本评估。通过将SigmaRL训练的策略[2]部署于所有三个领域，我们揭示了性能退化的两类互补来源：仿真与硬件控制栈之间的架构差异，以及环境真实度提高引发的仿真到现实差距。开放源代码设置使得在真实、可复现条件下系统分析MARL的仿真到现实挑战成为可能。
原文摘要 · Abstract (English)
We present a reproducible benchmark for evaluating sim-to-real transfer of Multi-Agent Reinforcement Learning (MARL) policies for Connected and Automated Vehicles (CAVs). The platform, based on the Cyber-Physical Mobility Lab (CPM Lab) [1], integrates simulation, a high-fidelity digital twin, and a physical testbed, enabling structured zero-shot evaluation of MARL motion-planning policies. We demonstrate its use by deploying a SigmaRL-trained policy [2] across all three domains, revealing two complementary sources of performance degradation: architectural differences between simulation and hardware control stacks, and the sim-to-real gap induced by increasing environmental realism. The open-source setup enables systematic analysis of sim-to-real challenges in MARL under realistic, reproducible conditions.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。