用星敏感器和地平仪测量星光与地平之间的“星光仰角”为观测量, 利用推广卡尔曼滤波方法实时估计航天器的最佳位置, 使航天器在失去地面遥控的情况下, 能够自主准确地确定运行轨道。由于航天器自主定位系统在工作期间会受到硬、软件等诸多因素的影响, 因而使其定位精度达不到预计要求。在此我们通过大量仿真计算, 指出一些对自主定位系统精度影响较大的因素, 并对它们进行了比较分析。 关键词: 自主式导航; 地平仪; 光学传感器; 方法; 精度 Abstract: The best po sit ion of spacecraft is real2t imely determ ined by using the extended kalman filtering in o r2 der that spacecraft can run in scheduled o rbit autonomously and accurately w ithout the remo te cont ro l of ground command. The elevat ion angular of star is observed by star senso r and ho rizontal inst rument. During thewo rk peri2 od of spacecraft, the system can be influenced by many facto rs related to hardw ere and softw ere, the p recision of o rbit determ inat ion can be sho rt of request. By emulat ing calculat ion, some facto rs that influence largely on the p re2 cision of determ ining po sit ion are po inted out, compared and analyzed earnest ly. Key words: A utonomous navigat ion; Ho rizon; L igh t sensit ive senso r; M ethod; P recision