【作者】 冷猛；

【导师】 李军；

【作者基本信息】 国防科学技术大学 ， 摄影测量与遥感， 2011， 硕士

【摘要】 利用星载传感器从太空对地球表面进行探测而获取信息,已广泛应用于社会、经济建设和国防建设等各领域。用户越来越期望对地观测卫星能够完成更加复杂多样的观测需求。对地观测需求复杂性主要体现在空域覆盖要求广、时域覆盖要求多次和频域覆盖要求多维等方面。复杂多样的对地观测需求给卫星任务规划带来了诸如规划算法复杂度高、规划效能低、规划结果用户满意度低、稀缺的卫星资源利用不合理等问题。对地观测需求分析是卫星任务规划的前端,通过对原始观测需求的各项特征要素进行分析,合理融合关联冗余的对地观测需求,优化多星接力观测流程,能有效提升卫星任务规划调度效能,使得观测结果更加适应不同类型的用户要求。论文针对用户提请的对地观测需求复杂性问题,引出了对地观测需求关联融合分析和多星接力对地观测需求流程分析。论文主要工作包括:1.深入研究了对地观测需求的各项特征要素,并根据对地观测需求的特征要素建立了对地观测需求描述模型;其次,根据对地观测卫星的相关特征属性及其搭载的对地观测传感器类型,建立了卫星对地观测能力描述模型;然后,阐述了用户提请的对地观测需求复杂性和卫星任务规划过程中的过载规划问题,从而明确了本文的研究重点,也即需要对用户提请的对地观测需求做关联融合分析和多星接力观测流程分析。2.研究了对地观测需求之间的关联性问题。本文引入关联测度定量化研究对地观测需求间的关联度,包括空域关联测度、时域关联测度和频域关联测度以及总的关联测度。分别研究了点目标与点目标的关联融合、区域目标与区域目标的关联融合和区域目标与点目标的关联融合,并建立了对地观测需求关联融合分析模型,针对此模型,用基于矩阵的层次融合分析方法对模型进行求解,然后通过实验验证了本文提出的方法,通过融合存在正关联关系的对地观测需求,可以有效使用对地观测卫星资源。3.研究分析了多星接力观测需求及其流程。本文针对典型的多星接力观测需求——区域目标覆盖探测及其关键部分监测,深入分析其组织实施观测过程,根据历史的观测信息,提出了覆盖探测和详情监测两步骤。当具备历史观测信息且能预测出区域目标中的关键部分的状态时,采取关键部分详情监测策略,否则采取覆盖探测策略。通过深入研究Petri网的相关理论以及运行机理后,建立基于着色Petri网(CPN)的多星接力观测流程层次化模型。构建的CPN层次化模型包含两层,第一层为多星接力观测层,第二层为观测方案生成层和观测数据传输层。最后通过遍历树算法和仿真方法对流程模型进行了逻辑正确性分析。最后,提出了基于启发式规则的区域目标覆盖探测及其关键部分监测多星接力观测方案推理方法。基于上述研究成果,设计并实现了具有良好人机互动的对地观测需求分析原型系统,验证和评估各项技术的有效性和实用性。更多还原

【Abstract】 That Earth Observing Satellites (EOSs) acquire information of the earth’s surface from outer space by using satellite sensors are widely used not only in the social, economic but also many other applications especially such as militay.That users rely on remote satellite for acquiring information has been accelerating, and makes earth observing request more complicated than ever before. The complexity of earth observing requests mainly are: wide coverage in geography,multi-time observeing and multi-fequency detecting. These complexities have bring obstacle for satellite task scheduling such as accelerating complexity of planning, degrading efficiency, low satisfaction of users and unrational use of satellllites and etc. Analysis for earth observing request is pre-part of satellite task scheduling. Both satellite task scheduling efficiency and users’satisfaction can be improved through earth observing request analysis.On the basis of the complexity of earth observing requests, analysis of cohesion and agglomeration for earth observing request was stated and flow model for multi-satellite relay observing was constructed.The main work of this dissertation can be concluded as the following three parts:1. Congsidering the characteristics of observing requests and remote satellites, two models were constructed to describe the earth observing requests and the capability of satellites respectively.The complexity of earth observing requests was listed. After that, two problems, cohesion between earth observing requests and relay observing flow for multi-satellits, were brought out.2. Through analyzing the cohesion between different requests proposed by huge mount of users, coherency measures was used to quantitate the cohesion of each two different earth observing requests. Coherency measures include time coherency, space coherency and frequency coherenc。And then agglomeration between two region-objects or between two point-objects or between point-object and region-object was considered. Later a model for analysis of cohesion and agglomeration for earth observing requests was established. And a hierarchical agglomerative algorithm based on matrix was proposed. Some experiments were implemented to valid correctness and practicability of our method.3. Facing representative multi-satellite relay observing problem, regional object detection and key part surveillance, a flow model based on colored petri net (CPN) was developed after study the theory of Petri Net. Finally, logic of this CPN model was verified based on coverability tree and by simulation test. Finally a reasoning method for multi-satellite observing schema was proposed based on heuristic rules.According to the research works presented above, an experiment system, which has friendly multi-view interfaces of interaction between human and machines, for earth observing request analysis, was designed and implemented as the platform to verify our research achievements.更多还原