惯性/天文角度组合导航在应用于高动态飞行器时,动态飞行环境变更会导致星光角度观测量发生程度不等的偏差,使得常规组合滤波方法误差显著增大.为此,本文提出了基于P值映射的观测质量自主评估及自适应滤波方法,并应用于惯性/天文角度组合导航系统.该方法根据历年可见导航星情况分解冗余观测子集,再由P值度量其含有观测量偏差的显著性水平.在此基础上,通过遍历每颗导航星所隶属子集得到其观测量质量值,最后对惯性/天文角度组合滤波增益进行自适应调节.仿真结果表明,本文方法能够实现天文高度角观测质量的自主在线评估,有效提高星光观测质量下降情况下惯性/天文角度组合导航的精度和适应性.
The integration of inertial navigation system (INS) and celestial angle navigation system are affected by flight conditions. When they are applied to highly dynamical aircrafts, there will appear biases in starlight observations as well as additional navigation errors for conventional integrated filtering approach. To solve those problems, we propose a novel P-value mapping-based quality evaluation for observations and introduce the adaptive filtering approach in the INS/celestial angle integrated navigation system. The observed stars are divided into redundant subsets; and the bias level of each subset is evaluated by a P-value. The observation quality of each star is determined based on the P-values of subsets which the star belongs to, and adaptively applied to adjust the gain of INS/celestial angle integrated filter. Simulation results show that the proposed approach can evaluate observation quality online autonomously, and effectively improve the precision and adaptability of the INS/celestial angle integrated navigation system when the starlight observation quality is declined.