Dynamic programming algorithm for satellite orbit task merging problem

Xiaolu Liu,Baocun Bai,Yingwu Chen,Jufang Li

2009-01-01

Abstract:Remote sensing satellites are platforms equipped with imaging instruments that orbit the Earth to take photographs of specific areas. To get required photographs each target is associated with a special time window and a given look angle. Due to the constraint of satellite's field of view (fov), a satellite has to roll its camera to get target images. However, some remote sensing satellites have rigid constraints on slew activities. Take HJ1A Satellite as an example, it can slew four times at most in a single orbit. That means the satellite can image no more than four times by different look angles, which greatly limits the satellite's capability. When multiple targets located nearby, or the intervals between their corresponding observations are not long and differences between their best look angles are not prominent, they might be observed together in one swath by the same look angle. Therefore, Satellite Orbit Task Merging Problem (SOTMP) is proposed. A dynamic programming approach is developed to find global optimal solution for this problem. The algorithm includes two search processes: forward and backward search. This bi-orientation search mechanism can ensure an optimal solution. Finally, a computational experiment is executed to test the approach. And the results justify the validness and correctness of the algorithm.