Performance-Dominant Fly-Around Orbit Design for Potential Pursuit-Evasion Game

This paper presents a performance-dominant fly-around orbit design methodology involving potential pursuit-evasion scenarios, focusing on enhancing navigation performance and maximizing pursuit success rate. Specifically, the navigation performance is quantified using the Fisher Information Matrix (FIM) derived via maximum likelihood estimation, while pursuit success is estimated via a Monte Carlo Tree Search (MCTS) algorithm. To improve computational efficiency, backpropagation (BP) neural networks are trained as surrogate models to enable rapid prediction of these performance metrics. Key mission constraints such as solar illumination and relative distance are incorporated to ensure payload performance. By integrating performance evaluation with constraint modeling, the proposed method enables the generation of optimized fly-around orbits, where mission constraints are incorporated into the overall objective via a penalty function mechanism. Simulation results show that the proposed method achieves orbit optimization and enables efficient, interpretable strategy design for pursuit-evasion scenarios.