<p>Basic properties of AUS demo.</p>

Enhancing the resilience of Autonomous Unmanned Swarms (AUS) requires policies that remain effective under severe, structured disruptions while respecting the heterogeneous semantics of inter–subsystem interactions. Existing reinforcement learning (RL) approaches typically aggregate first–order neighborhoods in a path–agnostic manner, thereby blurring typed, ordered, and directed multi–hop dependencies encoded by domain meta–paths. We propose MPGPD-RC, a Meta- Path Guided Policy Distillation framework for Resilient Coordination that couples: (i) meta-path–guided embeddings learned by path-specific graph attention with contrastive reconstruction and attention fusion, and (ii) a teacher–student scheme in which a PPO teacher trained with a relaxed meta-path mask provides trajectories, and a student aligns both action distributions (KL) and trajectory-level structural codes via path-aware contrastive learning. Empirical evaluations validate that MPGPD-RC consistently surpasses state-of-the-art baselines across diverse perturbation scenarios by modeling complex, high-order dependencies that underpin resilient coordination.

提升自主无人集群 (Autonomous Unmanned Swarms, AUS) 的弹性，需要设计既能在严重结构化扰动下仍保持有效，又能兼顾子系统间交互异质语义的策略。现有强化学习 (reinforcement learning, RL) 方法通常以路径无关的方式聚合一阶邻域，从而模糊了由领域元路径编码的类型化、有序且有向的多跳依赖关系。本文提出MPGPD-RC，即面向弹性协同的元路径引导策略蒸馏框架（Meta-Path Guided Policy Distillation framework for Resilient Coordination），其耦合了两项核心设计：(i) 结合对比重构与注意力融合的路径专属图注意力所学习的元路径引导嵌入；(ii) 师生协同架构：其中采用松弛元路径掩码训练的PPO (Proximal Policy Optimization) 教师模型生成轨迹，学生模型则通过路径感知对比学习对齐动作分布（KL散度）与轨迹级结构编码。实验评估结果表明，MPGPD-RC通过建模支撑弹性协同的复杂高阶依赖关系，在各类扰动场景中均能持续优于当前最优基线模型。