Satellite relay tracking of mobile targets holds significance in military reconnaissance and civilian monitoring. However, traditional methods that rely on preplanned observation sequences struggle with autonomous real-time handovers in dynamic scenarios, often causing tracking interruptions or loss of efficiency. This paper proposes an STGNN-based visibility prediction method for collaborative satellite relay planning to achieve real-time trans-satellite task transfer. First, we model satellite-target spatiotemporal correlations via STGNN to predict visibility, achieving a prediction accuracy of greater than 95% that meets realtime decision requirements. Subsequently, a collaborative handover mechanism dynamically aggregates adjacent satellite visibility predictions and status parameters, screening optimal successors through relay selection to ensure tracking continuity. Simulations demonstrate prolonged target tracking and validate effectiveness in dynamic environments, offering a new technical pathway for autonomous satellite cluster coordination.
扫码关注我们
求助内容:
应助结果提醒方式:
