组网成像卫星任务规划鲁棒性建模与算法研究

发布时间：2018-01-07 17:15

  本文关键词：组网成像卫星任务规划鲁棒性建模与算法研究　出处：《哈尔滨工业大学》2016年硕士论文　论文类型：学位论文

  更多相关文章： 成像卫星 任务规划 证据理论 鲁棒性模型 改进遗传算法

【摘要】：组网成像任务规划呈现出卫星组网规模大、成像要求高、时效性强、资源种类多等特点。成像卫星任务规划受到目标插入、取消、属性变动、气象因素变动等风险要素影响的程度越来越广。单纯的静态环境下的规划方案已经不能满足多变环境中对方案稳健性的要求。在此背景下,编制具备较强鲁棒性的成像卫星任务规划方案需求正是本文研究的前提和基础,也是资源调度,方案规划的现实需要。本文为解决组网成像卫星任务鲁棒性规划而开展的具体的工作内容如下:(1)依据成像卫星任务执行的过程,分析、归纳了不确定性要素,以及根据鲁棒性需求采取不同的应对策略。根据各种风险要素的特点,利用证据理论综合信息来处理影响成像卫星任务执行的不确定性。同时,寻求新的鲁棒性指标,以便在模型求解过程中获取具备较强鲁棒性的执行序列。(2)根据组网成像卫星任务规划现实要求,结合风险要素构建鲁棒性调度模型。在鲁棒性规划模型构建进程中,进行符号化表示和解释。根据研究问题的重点,构建满足限制条件的鲁棒性模型和优化函数。(3)针对模型求解型NP-Hard问题,本文应用以遗传算法为基础对模型求解。新算法以遗传算法为基础,结合模拟退火算法。新方法要应用新的问题染色体组成,避免遗传算法只取局部最优解的问题,以及应用更加高效的进化策略。(4)模型仿真案例实验。本文利用仿真软件STK工具包设立相应的场景。规划场景中包含若干成像卫星和分布均匀的地面目标,根据构建的鲁棒性模型和模型求解算法求解场景规划。案例仿真算例得到规划方案,以此进行分析。
[Abstract]:The imaging mission planning of the imaging network shows the characteristics of large scale of satellite network, high imaging requirements, strong timeliness, and many kinds of resources. The mission planning of imaging satellite is subject to target insertion, cancellation and attribute change. The degree of influence of risk factors such as the change of meteorological factors is more and more extensive. The planning scheme under the simple static environment can not meet the requirements of the robustness of the scheme in the changeable environment. In this context. It is the premise and foundation of this paper to develop a robust mission planning scheme for imaging satellite, and it is also the resource scheduling. The specific work of this paper to solve the robust planning of imaging satellite mission is as follows: 1) according to the imaging satellite mission implementation process, analysis. The uncertain factors are summarized, and different coping strategies are adopted according to the requirements of robustness, and according to the characteristics of various risk factors. The evidence theory is used to synthesize information to deal with the uncertainty that affects the performance of the imaging satellite mission. At the same time, a new robust index is sought. In order to obtain the execution sequence with strong robustness in the process of model solving.) according to the practical requirements of mission planning of networked imaging satellite. The robust scheduling model is constructed by combining the risk factors. In the process of constructing the robust planning model, the symbolic representation and interpretation are carried out. According to the focus of the research. To solve the NP-Hard problem of model solving, this paper applies the genetic algorithm to solve the model. The new algorithm is based on the genetic algorithm. Combined with simulated annealing algorithm, the new method should apply new problem chromosome composition to avoid the problem that genetic algorithm only takes local optimal solution. And applying a more efficient evolutionary strategy. Model simulation case experiment. This paper uses the simulation software STK toolkit to set up the corresponding scene. The planning scenario includes several imaging satellites and even distributed ground targets. According to the robust model and the algorithm of solving the model, the scenario planning is solved, and the case simulation example is used to get the planning scheme.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：V474
，

本文编号：1393496