面向博弈任务的空间飞行器集群高动态构型 控制方法

A dual-layer semi-decoupled collaborative control architecture is proposed to address high dynamic configuration changes during spacecraft cluster and target spacecraft games. The problem is decomposed into an inner layer “orbital game between navigation spacecraft and target spacecraft” and an outer layer “collaborative control of spacecraft clusters”. For the inner layer, a unified control model is established for three typical configurations— collaborative surrounding, semi-encirclement blocking, and linear barrier—with a zero-sum game strategy designed and nonlinear terms solved rapidly using composite neural networks. For the outer layer, a cluster game optimization model is established, and a distributed estimator is designed to solve the control strategy for accompanying spacecraft. Simulation results demonstrate that the proposed architecture enables spacecraft clusters to achieve specified configurations according to target maneuvering characteristics while ensuring stable control and dynamic response, validating its rapidity, real-time performance, and robustness.