Optimal Scheduling of $K\text{-}M\text{-}N$ GEO On-Orbit Service Network Based on Hybrid Simulated Annealing Algorithm

A $K\text{-}M\text{-}N$ on-orbit service network architecture and its mission scheduling problem are proposed and studied, in which $K$ space service stations and $M$ service spacecraft are scheduled to service for $N$ target spacecraft. The service spacecraft, which are initially docked with the space service station in the geosynchronous region, set off through orbital maneuver to conduct servicing operation for one or multiple target spacecraft upon receiving the service command. A mission scheduling strategy of on-orbit service for the target GEO spacecraft is proposed based on space service stations, in which the service spacecraft are scheduled to serve all target spacecraft with the lowest cost of fuel for the orbital transfers. As the mission scheduling for such a $K\text{-}M\text{-}N$ GEO on-orbit service network is NP-hard, a hybrid simulated annealing optimization algorithm is put forward based on a mechanism of dynamic reverse processing of encoding and decoding to solve this problem. Four numerical case studies are presented and their results are compared with those from the literature to demonstrate the effectiveness of the approach. A series of different network model simulations and Mann-Kendall tests prove that multispace service stations can significantly reduce total fuel consumption for transfers, and the initial layout is also very important. Besides, the initial distribution of the service spacecraft also has an impact on total fuel. However, the quantitative impact of the service spacecraft is uncertain and needed to be analyzed on a specific basis, sometimes increasing the quantity of service spacecraft can reduce the total fuel, sometimes it has no effect, sometimes it may even cause negative effect when the number of service spacecraft is too large and unreasonable.

On-orbit servicing (OOS), refers to the execution of space operations through the exterior vehicle activities of astronauts, robots or both to extend the mission life of a spacecraft or improve the ability of a spacecraft to perform on-orbit missions. With the development of relevant enabling technologies, OOS will become a business which can be implemented commercially and in a large scale, in which one or multiple space service stations, as the space resource storage warehouse of the satellites in orbit, run in the earth orbit for a long time, and the service spacecraft will provide service to the target spacecraft in turn after completing the rendezvous and docking. In the process of servicing, the service spacecraft will return to the space service station for the replenishment when its resource is insufficient. In this paper, on-orbit service network scheduling based on space service station is studied, which is NP-hard. A hybrid simulated annealing optimization algorithm is put forward based on a mechanism of dynamic reverse processing of encoding and decoding to solve this problem. The researches in this paper are of great significance to the scale design and scheduling of on-orbit services in the future, meanwhile have a certain appeal for the relevant operational research scholars.