Optimal Mission Planning for Multiple Agile Satellites Using Modified Dynamic Programming

Abstract

This paper aims to optimize imaging mission planning for multiple agile Earth observation satellites using a modified dynamic programming (MDP) algorithm. The mission planning problem is mathematically modeled using mixed-integer linear programming, which defines the objective function and constraints. An MDP algorithm is proposed to address the challenge of extended computation time inherent in typical deterministic dynamic programming, aiming to optimize the imaging scheduling problem. Numerical simulations are conducted under various target numbers and satellites in a relatively narrow mission area with a high target density. The simulation results illustrate the stage configuration and target observation sequence over time, and this paper investigates the achievement rate, measured as the ratio of obtained profit and target observations to the total. The trend of computation time is also analyzed across all mission scenarios. Ultimately, the validity of the MDP algorithm is confirmed, as it demonstrates superior performance compared to the genetic algorithm, the branch-and-bound method, and the greedy algorithm.