Spatial-temporal intention representation with multi-agent reinforcement learning for unmanned surface vehicles strategies learning in asset guarding task

Yang Li, Shaorong Xie, Hang Yu, Han Zhang, Zhenyu Zhang, Xiangfeng Luo

Abstract: Highlights•Learning multi-USV asset guarding strategies under dynamic changes in adversary intention.•Reasoning intentions by mapping behavioral features to intention types using prior rules.•Representing intentions in both spatial and temporal dimensions with the spatial–temporal attention network.•Intention representation combined with MARL to train multi-USV strategies.•Validating the method in various virtual asset guarding environments.