Request-Centric Modeling and Architecture Optimization for Earth–Moon Space Logistics

This work presents a campaign-level optimization framework for designing space transportation architectures to support servicing operations in cislunar orbit. The goal is to service multiple client satellites—requiring refueling, inspection, and repair—using a fleet of high- and low-thrust spacecraft. A multi-objective optimization approach determines the optimal number and type of servicing spacecraft, their trajectories, and orbital depot locations, aiming to maximize the number of fulfilled servicing requests while minimizing both fuel consumption and task completion time. The cislunar region is represented by six orbital nodes: low Earth orbit, geostationary transfer orbit, geostationary Earth orbit, Earth–moon L1 and L2 halo orbits, and low lunar orbit. To manage the computational complexity, a two-step approach is adopted. First, transfer trajectory costs between orbital nodes are computed for both high-thrust and low-thrust servicing spacecraft using four transfer strategies: i) bi-impulsive minimum-Δ⁢𝑉 direct transfers, ii) flyby-supported minimum-Δ⁢𝑉 transfers, iii) minimum-time direct transfers, and iv) invariant manifold transfers. Second, these costs are integrated into a two-level heuristic optimization algorithm with embedded feasibility constraints. The resulting Pareto-optimal front highlights tradeoffs among feasibility, affordability, and effectiveness of multiple potential solutions. The methodology is tested on two one-year servicing mission scenarios involving three client satellite constellations.