Robust Trajectory Design for Spacecraft Rendezvous Considering Navigation Performance and Passive Safety

This study proposes a robust trajectory design method for spacecraft rendezvous missions under angles-only navigation conditions. The method simultaneously considers estimation accuracy and passive safety by integrating covariance dynamics into a stochastic differential dynamic programming (SDDP) framework. By explicitly propagating the estimation error covariance within the optimization, the approach can generate trajectories that actively improve observability while satisfying passive safety constraints, a trade-off that has not been explicitly addressed in previous rendezvous studies. Numerical examples demonstrate that both fuel-oriented and observability-aware solutions can be systematically obtained by adjusting the weighting parameters in the cost function, revealing a clear trade-off among fuel consumption, terminal estimation accuracy, and passive safety margin.