基于预测跟踪星表提高星敏感器实时性的方法

为提高星敏感器在星跟踪失败时向全天识别模式转换的效率，提出了一种在星跟踪的同时，并行预测跟踪导航子星表的方法，以随时做好工作模式切换的准备，提高星敏感器实时性。设计了分层存储、逐层索引结构的导航星表，实现了均匀化筛选导航星、构建导航子星表及对数据快速搜索和处理；由飞行态数据预测下一时刻星敏感态，并通过星敏感器视轴指向与子星表地址之间的索引关系，预测跟踪子星表的地址；星跟踪失败时，全天识别在4个冗余子星表中完成，简化识别过程，提高了识别效率。实验结果表明，设计的导航星表较为科学、合理，预测跟踪星表方法具有较强的鲁棒性，在高动态条件下有效提高了星敏感器模式切换的实时性。

Method of improving real-time of star sensor based on predicting and tracking star catalog

To improve the efficiency of star sensor during the transition from star tracking mode to full-sky star identification mode, a novel method of synchronously predicting and tracking guidance star sub-catalog while in star tracking mode was proposed. The work mode transition was always prepared, therefore the real-time property of star sensor was improved. The overall catalog architecture which was hierarchical storage and sequentially mapping' was designed. Based on this architecture, guidance stars were homogenized, the overlapped sub-catalogs were generated, and fast searching and management of data were realized. The attitude of star sensor at next time was predicted by using the attitude data of air vehicle. The address of sub-catalog was predicted and tracked by the built index relationship between bore sight of star sensor and addresses of sub-catalogs. Once star tracking failed, full-sky star identification implemented in four redundant sub-catalogs, therefore the process of star identification was simplified, and the efficiency was improved. Experimental results show that the designed star catalog is more scientific and reasonable. The proposed method has strong robustness. The real-time property in work mode transition of star sensor at highly dynamic condition can be effectively improved.