Energy-Dependent Mission Planning for Agile Earth Observation Satellite

Abstract

This paper investigates the energy-dependent mission planning of an agile earth observation satellite. According to the similarities between energy-dependent mission planning and the dynamic traveling salesman problem (DTSP), the energy-dependent mission planning is converted into a DTSP through two mappings from observation angle and energy to city and distance. In addition, to eliminate the uncertainty of feasible attitude trajectories/energy for a given scheduling strategy, a multiple-stage optimal energy factor is developed for the model extension. The paper further uses the time-optimal minΔTtransi−1,i in the transition time constraint as an input of the model to enlarge the optional execution time for the subsequent target in a consecutive observation. To solve this problem, a hybrid method integrating the Gauss pseudospectral method and the genetic algorithm (GPM-GA) is proposed which uses the genetic algorithm to generate the feasible solutions for scheduling process, whereas the Gauss pseudospectral method (GPM) is used to optimize the energy and the transition time constraint parameter for each solution. Extensive simulation results show that, compared with the classical genetic algorithm (CGA), the energy consumption and simulation time of the proposed algorithm are decreased effectively. In particular, the simulation time decreases more obviously with larger target sizes. Furthermore, attitude trajectory projections of satellite motion provided by GPM-GA are much smoother. These numerical and visualization results demonstrate the superiority of GPM-GA in terms of energy efficiency, computational efficiency, and attitude trajectory smoothing.