Research on Autonomous Decision Method of Multi-unmanned System Based on Hierarchical Behavior Tree

To address the issues of slow and inflexible collaborative decision-making among multiple unmanned platforms in future cross-domain cooperative combat scenarios, this paper proposes an autonomous decision-making method based on hierarchical behavior trees, constructing a three-layer behavior tree model spanning “mission-cluster-platform”. By customizing multiple types of nodes, it enables top-down decomposition, dynamic allocation, and online on-the-spot adjustment of complex tasks. In a cross-domain anti-submarine search simulation experiment involving multiple unmanned platforms, it has achieved fully autonomous collaborative combat capable of commanding no less than 50 heterogeneous unmanned platforms of three types, with an online generation delay of platform action instructions of only 250 ms. The hierarchical behavior tree architecture proposed in this paper supports node reuse and flexible configuration, significantly reducing command load and providing scalable technical support for fully leveraging the multi-domain dynamic complementary advantages of multi-unmanned cluster systems.