自适应加权质心算法在高精度星点定位中的应用
发布时间：2018-12-18 18:09
【摘要】：为实现空间飞行器在高目标星等、像移和光学系统像差等因素影响下的高精度星点定位,提出了一种自适应加权质心细分定位算法。基于最大似然估计推导、建立数学模型并进行仿真实验。结果表明,与质心法、加权质心法及高斯拟合法等传统星点定位算法相比,提出的算法定位精度提高了50%以上,且具有良好的收敛速度,经过5次迭代后,质心定位误差相对较稳定,可以满足实际应用中对实时性的要求。
[Abstract]:An adaptive weighted centroid subdivision localization algorithm is proposed to achieve high precision star location under the influence of high target magnitude, image shift and optical system aberration. Based on the derivation of maximum likelihood estimation, the mathematical model is established and the simulation experiment is carried out. The results show that compared with the traditional star location algorithms such as centroid method, weighted centroid method and Gao Si fitting method, the proposed algorithm can improve the accuracy of star location by more than 50%, and has a good convergence rate. After five iterations, the proposed algorithm has a good convergence rate. The centroid positioning error is relatively stable, which can meet the real-time requirements in practical applications.
【作者单位】： 中国科学院长春光学精密机械与物理研究所;
【基金】：国家自然科学基金(61205143)
【分类号】：V448.2
，
本文编号：2386274
[Abstract]:An adaptive weighted centroid subdivision localization algorithm is proposed to achieve high precision star location under the influence of high target magnitude, image shift and optical system aberration. Based on the derivation of maximum likelihood estimation, the mathematical model is established and the simulation experiment is carried out. The results show that compared with the traditional star location algorithms such as centroid method, weighted centroid method and Gao Si fitting method, the proposed algorithm can improve the accuracy of star location by more than 50%, and has a good convergence rate. After five iterations, the proposed algorithm has a good convergence rate. The centroid positioning error is relatively stable, which can meet the real-time requirements in practical applications.
【作者单位】： 中国科学院长春光学精密机械与物理研究所;
【基金】：国家自然科学基金(61205143)
【分类号】：V448.2
，
本文编号：2386274
本文链接：https://www.wllwen.com/kejilunwen/hangkongsky/2386274.html
