IMPLEMENTING INTEGER LINEAR PROGRAMMING TO OPTIMIZE ON-ORBIT SERVICING MISSIONS

Historically, satellite maintenance has been conducted by manned missions launched to service only one satellite. This carries with it inherent risks and large costs. On-orbit servicing (OOS) is a mission concept for employing satellite servicer vehicles, either employed from an orbital “depot” or launched with mission-specific supplies, to conduct maintenance of satellites in orbit. In order to service multiple satellites in one mission, these servicer vehicles must be capable of completing multiple sequential orbital transfers, which are costly in terms of both fuel and time. With limited capacity for fuel onboard, selecting the proper sequence to minimize the fuel or time spent on each mission is of great interest. This thesis formulates an integer linear program for determining the optimal sequence of satellites to be visited. Our model accounts for changes to satellite orbits over time and allows for the possibility that the servicer may wait between transits in order for the orbits to reach a more desirable configuration. We implement the model in Python and find solutions to a series of satellite sets of various sizes to examine the model’s computation time and the fuel usage and time required to execute the resulting sequence of transits.