Answers are grounded in this paper and cite their sources.
On-orbit inspection is of growing importance as low-earth orbits become more cluttered, and new tools are needed to facilitate satellite maintenance and repair. We presume that an inspecting satellite conducts a fly-by maneuver around a torque-free, tumbling satellite, with uncertain initialization and a specific key point of interest on the target. However, the inspecting satellite is limited in what it can observe, due to field-of-view and occlusion constraints. Further, we presume that sensing quality is dependent on how close the satellite is to being directly overhead of the key points of interest on the inspection target. Finding feasible trajectories for the on-orbit inspection is difficult, not only because of the nonlinear quaternion kinematics and nonlinear constraints, but also because of the non-Gaussian stochasticity introduced through the uncertainty of the initial condition of the tumbling satellite. We seek to identify feasible trajectories for inspection, and to explore trade-offs between sensing performance, deviations from the nominal fly-by trajectory, and fuel expenditures. We formulate this problem as a type of stochastic optimization problem, and leverage sample average approximation (SAA) to solve it approximately. Solutions provide a computationally efficient way to assess inspection trajectories that would otherwise be extremely difficult to synthesize. We demonstrate the efficacy of our approach, and examine trade-offs between the multiple elements of the cost function, on both a stable tumbling object, and on a chaotic tumbling object.
No takes yet. Share an insight, caveat, or question.
Sivaramakrishnan et al. (2025) studied this question.