Multi-agent airspace hierarchical interception decision model based on area coverage

The rise of low-attitude economy has posed threats to critical infrastructure. Currently, there is a lack of research on task decision-making models for aerial target interception by low-speed agent swarms with limited interception capability. Therefore, this paper proposes a multi-agent airspace hierarchical interception decision model based on area coverage optimization for critical infrastructure protection. To quantify the collective interception efficiency, we first establish a joint interception probability model for multi-agent swarms that accounts for target distribution. Subsequently, we employ the Elite Driven Gray Wolf Optimizer (EDGWO) to optimize the positions of agents in the target prediction zone, thereby enhancing the joint interception probability. Finally, under the assumption of mutual independence among multiple interception layers, we propose a temporal evolution mechanism for multi-agent hierarchical interception tasks, based on which a hierarchical interception decision model is constructed and solved via the discrete EDGWO algorithm. Simulation experiments verify that, compared with manually configured baseline under the same constraints, our model achieves a slight but meaningful 0.67% improvement in area coverage and a 2.8% increase in the joint interception probability. The hierarchical interception decision model yields an optimal 3-layer interception strategy achieving the expected 95% successful interception probability, with 7, 5, and 4 agents deployed in each layer, respectively. This research offers practical guidance for low-altitude air defense of urban critical infrastructure using low-performance agents.