Abstract: To address the real-time target assignment problem in heterogeneous missile swarms cooperatively intercepting highly maneuverable UAV swarms, an event-driven, dynamic-clustering-based target assignment algorithm is proposed. By jointly accounting for key factors such as strong target maneuverability and the limited communication range of missiles, an integrated framework combining “event-driven control, dynamic clustering, and multi-level auction with conflict resolution” is established. Within this framework, to cope with highly maneuverable targets and their dynamically changing spatial distribution, a novel swarm-feature-based event-triggering mechanism is designed, which initiates reassignment only when the interception scenario changes significantly, thereby avoiding the communication redundancy and scheduling latency caused by periodic reassignment. To address the constraint of limited communication range, an adaptive dynamic clustering scheme based on K-means++ is introduced, together with a corresponding multi-level auction and conflict-resolution mechanism, enabling low-overhead and highly robust cooperative target assignment in unbalanced many-to-one engagement scenarios. Simulation results demonstrate that the proposed method markedly reduces communication load and reassignment frequency, while achieving terminal-phase interception performance superior to that of benchmark distributed online assignment methods and close to that of a centralized online assignment approach.

Key words: heterogeneous missile swarms, event-driven, dynamic clustering, unbalanced assignment, target assignment