Abstract:

Intelligent unmanned aerial vehicle swarm combat exhibits multi-scale characteristics of micro-level tactical autonomy, meso-level resource constraints, and macro-level network collaboration. In view of the problem that a single-scale modeling method is difficult to simultaneously characterize individual decision-making details, resource flow process, and system collaboration mechanism, a multi-scale modeling method integrating multi-agent modeling, system dynamics, and complex network theory was proposed. By designing six formal coupling operators to achieve cross-layer information mapping, a time progression mechanism based on a master-slave clock and hybrid synchronization was established, and a proof of error boundedness was provided. A prototype system was implemented on a simulation platform, and ablation and tactical emergence experiments were conducted through a red-blue unmanned aerial vehicle swarm confrontation scenario. The results confirm that the multi-scale modeling method has significant advantages over the single-scale method, which provides a feasible engineering implementation path for the modeling of intelligent unmanned aerial vehicle swarm combat systems.

Key words: intelligent unmanned aerial vehicle swarm, multi-scale modeling, multi-agent, system dynamics, complex network, tactical emergence