应用近似算法的光学遥感卫星区域目标成像任务规划方法

Region Target Imaging Task Scheduling Method Based on Approximation Algorithm for Optical Remote Sensing Satellite

光学遥感卫星成像任务规划的研究对象多以点目标为主,对于区域目标的处理,多是将其转化为点目标后求解,因此存在精度不高和效率过低等问题。同时,还存在对于长时间周期的规划问题耗时过长,以及大区域长时间任务规划性能较差等问题。文章根据卫星成像任务规划的约束条件,建立成像任务规划模型,分析并归纳光学遥感卫星轨道及星下点轨迹的特点,认为星下点轨迹在中低纬度区域的投影可近似认为是直线;通过将区域目标向直线投影,将二维平面区域转化为一维线状区域,对模型进行求解。该方法能大幅降低求解区域目标问题的复杂度,在一维线状区域的基础上对侧摆进行处理,同时能够提高长时间周期规划的性能。利用环境减灾-1A(HJ-1A)卫星进行验证,结果表明,文章提出的方法能够在较短的时间内解决大区域长时间成像任务规划问题。

Most of the imaging task scheduling study objects of optical remote sensing satellite are based on point target. For region target, most of the researches convert them to point target, which has disadvantages of low-precision and low-efficiency. Additionally, solving the problems of long-period scheduling is time-consuming, and large-area and long-period imaging task schedu- ling is compared with poor performance. In this paper, imaging task scheduling model is built ac- cording to constrain conditions, and optical remote sensing satellite orbit and the track of sub-sat- ellite points are induced and analyzed. The projection of track of sub-satellite points is approxi- mately considered as straight line. The model is solved through region target projected to a line, two-dimension surface region converted to one-dimension linear region. The complexity of solvingregion target problems is reduced significantly. The side swing is handled in one-dimension linear region, and the performance of solving long-period scheduling problem is improved at the same time. The verification results of HJ-1A satellite show that method presented in this paper can solve the large-area and long-period imaging task scheduling problem in a short time.