To achieve space-based optical surveillance, bi-satellite geometric positioning model was established. Accord-ing the binary geometric positioning theory, with angular-measure-only principle, the position of the target was obtained by using least square method. Because the result of star chart processing would directly affect the final positioning accura-cy, the grey scale characteristic was analyzed. Comparing several different star point extraction algorithms by using real chart, some extraction algorithms were validated and compared and two feasible point extraction algorithms were found out. The bi-satellite geometric positioning model was simulated. The algorithm positioning error was combined by using Monte Carlo method. The conclusions showed that the positioning error was less than 10-5m without error source. And the position-ing error was less than 100m when taking four main error sources into consideration and the attitude measurement error gave the greatest impact on the precision of positioning method. And the positioning error was approximately normal distri-bution and almost liner with error sources. So this positioning algorithm could locate the target with high accuracy. The re-sult of this paper has the value of engineering application.%为实现空间目标的天基光学观测,利用双星定位原理,采用仅测角的观测方式,结合最小二乘法建立了双星几何定位模型.考虑到星点提取作为光学定位算法中的重要一环将直接影响最终的定位精度,因而针对就星图处理部分,分析了星图的灰度特点,采用阈值分割法进行了去噪处理,选取了几种质心法对星点进行提取,利用真实星图对提取结果进行比较,选出了形心法和阈值质心法2种较为可行的星点提取算法.通过仿真计算和蒙特卡洛误差分析结果可知,不考虑误差时,定位误差小于10-5 m;考虑主要误差源时,定位误差小于100 m;且姿态误差对定位方法的精度影响最大,数值上可以达到总误差的0.9倍以上,速度误差的影响最小,只有总误差的1/30,可以忽略;定位误差近似呈高斯分布且误差的大小与误差源近似呈线性关系.由此可见,文章所建立的定位算法可以对空间目标实现较为精确的定位.
展开▼
