An optimization approach for designing optimal tracking campaigns for low-resources deep-space missions

This work contributes to the autonomous scheduling of orbit determination campaigns for tracking spacecraft in deep-space by developing a dedicated optimisation algorithm. Given a network of available ground stations, the developed method autonomously generates optimized tracking observation campaigns, in terms of stations to use and time of measurements, which minimize the uncertainty associated to the state of the satellite. The outcome is a set of optimal solutions characterized by different allocated budgets, among which the operators can choose the most appropriate or promising one. The developed approach relies on a Structured-Chromosome Genetic Algorithm that copes with mixed-discrete global optimization problems with variable-size design space. This operates on a hierarchical reformulation of the problem by means of revised genetic operators. The estimation of the spacecraft state and its uncertainty, given a set of measurements is performed using a sparse Gauss-Hermite Kalman Filter. The proposed approach has been tested to the design of observation campaigns for tracking a satellite in its interplanetary cruise to an asteroid. Uncertainty is considered in the initial conditions, execution errors and observation noises.