Analysis of Multi-Orbit Multi-Payload Injection Scenarios for an Upper Stage

In the recent years, the number of launches has increased dramatically, showing a tendency beyond the current space transport systems. Such a problem, partially provoked by the increased number of space users due to the industry inclusion, as well as by the shift of interest towards smaller satellites and constellation missions, demands of innovative and economical solutions. One of these is the capability of an upper stage to directly inject multiple satellites into their respective differentiated orbits, reducing this way the number of necessary launches while allowing the growth of the space environment usage. Such multi-payload multi-orbit injection trajectory requires of a control law that can provide the manoeuvres while minimising both the fuel consumption and the overall mission time. Its definition is not straight forward and requires solving a complex optimisation problem composed by the visiting sequence and the individual transfers. The current paper proposes a strategy to define such a trajectory by diving the problem into two: a preliminary bi-level bi-objective optimisation algorithm that determines the ideal orbit order and the approximate necessary Lambert transfers. The result is a flexible algorithm that can provide for the full set of transfers given a set of orbits and satellites to be delivered, regardless of the final injection orbits and mass properties of the payloads. This algorithm is then used to analyse a certain payload injection scenario, given realistic mission data from the space industry. The performance of the upper stage in terms of total time and consumed fuel is examined, providing an assessment on the feasibility of this type of mission. The same case scenario is then studied under different conditions in terms of randomness, decision criteria and routing constraints to achieve further understanding on the possible real case scenarios. It is shown that in all cases the algorithm is able to converge towards feasible solutions, and that the nature of the resulting trajectories depends highly on the scenarios themselves, but also on the logical operation of the bi-level optimisation loop.