GEO碎片天基观测图像速度估计与动态规划检测
发布时间：2018-11-17 06:57
【摘要】：本文提出了一种基于位置预测的动态规划方法,用于快速检测GEO带天基观测图像中的暗弱碎片。首先对实测数据进行分析,得出不同GEO碎片在图像中的运动速度相似,接着,进一步对数据做拟合,得到了GEO碎片速度和观测卫星星下点纬度之间的映射模型,并提出了GEO带天基观测图像中碎片速度估计方法。最后,利用估计速度预测碎片位置,在传统动态规划递归方程中采取位置信息加权,得到了基于位置预测的目标搜索范围,从而减少了递归方程中的目标状态个数。实测数据验证了速度映射模型的拟合偏差在1pixel以内。选取一个典型观测周期的实测数据进行实验,结果表明本文方法的检测时间比传统动态规划方法减少了90%以上,虚警率降低了5.9%以上,适合于GEO带天基观测图像中暗弱碎片的检测。
[Abstract]:In this paper, a dynamic programming method based on position prediction is proposed for fast detection of dark weak debris in space-based observation images of GEO. By analyzing the measured data, it is found that the velocity of different GEO fragments in the image is similar. Then, the mapping model between the velocity of GEO fragments and the latitude of the observation satellite is obtained by further fitting the data. A method for estimating the velocity of debris in GEO space-based observation images is proposed. Finally, the estimated velocity is used to predict the debris position, and the position information is weighted in the traditional dynamic programming recursive equation, and the target search range based on the location prediction is obtained, thus reducing the number of target states in the recursive equation. The measured data verify that the fitting deviation of the velocity mapping model is within 1pixel. The experimental results show that the detection time and false alarm rate of this method are more than 90% and 5.9% less than that of the traditional dynamic programming method. It is suitable for the detection of dark weak debris in GEO space-based observation images.
【作者单位】： 北京跟踪与通信技术研究所;
【基金】：国家高技术863计划资助项目(No.2015AA7088061)
【分类号】：TP391.41
本文编号：2336867
[Abstract]:In this paper, a dynamic programming method based on position prediction is proposed for fast detection of dark weak debris in space-based observation images of GEO. By analyzing the measured data, it is found that the velocity of different GEO fragments in the image is similar. Then, the mapping model between the velocity of GEO fragments and the latitude of the observation satellite is obtained by further fitting the data. A method for estimating the velocity of debris in GEO space-based observation images is proposed. Finally, the estimated velocity is used to predict the debris position, and the position information is weighted in the traditional dynamic programming recursive equation, and the target search range based on the location prediction is obtained, thus reducing the number of target states in the recursive equation. The measured data verify that the fitting deviation of the velocity mapping model is within 1pixel. The experimental results show that the detection time and false alarm rate of this method are more than 90% and 5.9% less than that of the traditional dynamic programming method. It is suitable for the detection of dark weak debris in GEO space-based observation images.
【作者单位】： 北京跟踪与通信技术研究所;
【基金】：国家高技术863计划资助项目(No.2015AA7088061)
【分类号】：TP391.41
【相似文献】
相关期刊论文 前1条
1 强胜;易东云;潘晓刚;;基于天基观测的空间目标变轨识别算法研究[J];系统仿真学报;2009年12期
相关博士学位论文 前1条
1 李冬;天基观测目标跟踪、定轨及网络路由算法研究[D];国防科学技术大学;2012年
相关硕士学位论文 前1条
1 徐伟;GEO目标的天基观测任务规划研究[D];国防科学技术大学;2014年
，本文编号：2336867
本文链接：https://www.wllwen.com/kejilunwen/ruanjiangongchenglunwen/2336867.html
