An optimization framework with improved auction-based initialization for highly constrained on-orbit servicing mission planning

Abstract

The on-orbit servicing (OOS) mission planning problem for multiple geosynchronous earth orbit (GSO) satellites is a critical issue due to its huge economic value. However, given the complexity of the space scenario and the subjective nature of the formulation process, the problem is highly constrained in actual use, and the performance of conventional optimization algorithms decreases due to fuel or mission deadline constraints. With the goal of enhancing solution quality and speeding up the solution process, this paper proposes a novel optimization framework. The problem is decomposed into three sub-problems and then solved using the proposed three-level algorithm to lessen its complexity. In addition, a new initialization method is embedded to improve global search capabilities. Notably, considering the high computation efficiency of the auction algorithm (AA), an improved version is designed to generate primary solutions, which introduces the grouping and reallocation mechanism. Then, several variations of the primary solutions are used as the initial variables in the optimization. Numerical simulations demonstrate that the proposed methodology has higher computational efficiency and convergence performance than conventional strategies.