A Three-Stage Hierarchical Optimization Framework for Operational Imaging Performance in ΔV-Constrained Earth Observation Constellations

Earth observation satellite constellations must ensure rapid and reliable coverage of critical areas under propulsion constraints and environmental variability. Existing studies typically treat constellation geometry, maneuver optimization, and operational assessment as separate problems, limiting understanding of the structural origin of performance constraints. This study proposes a three-stage hierarchical optimization framework that separates intrinsic geometric potential, reactive multi-objective maneuver adaptability, and stochastic operational realizability. In Stage 1, a genetic algorithm optimizes constellation geometry with respect to global success rate across multiple areas of interest. In Stage 2, NSGA-II-based maneuver optimization evaluates the trade-off between success rate, time to first access, and total $\mathsf{\Delta }V$ expenditure. In Stage 3, operational performance is assessed using the Operational Imaging Response Time (OIRT) metric under probabilistic environmental filtering. Simulation results show that constellation geometry is the primary structural determinant of performance within the analyzed configuration range, while maneuver optimization provides moderate but measurable response-time improvements. Environmental filtering reduces achievable coverage relative to purely geometric predictions. The proposed framework establishes a consistent methodology linking constellation synthesis, maneuver planning, and operational response evaluation for time-critical Earth observation missions.