Abstract: The world′s first dedicated pulsar navigation test satellite, X-ray pulsar-based navigation-1(XPNAV-1), was successfully launched by China in November, 2016. The satellite is operating normally in orbit, and a large amount of observation data has been acquired. The measured data of a single pulsar was used to verify the effectiveness of the satellite orbit improvement and the pulsar navigation system through the geometric constraint method. However, the longterm orbit determination process using a single pulsar had divergence problems. Based on the XPNAV-1 satellite expansion test mission and the need for subsequent development of pulsar navigation, the observation data of multiple pulsars was used to study the autonomous orbit determination algorithm of satellites based on extended Kalman filter (EKF). Firstly, the orbital mechanics model and observation equations of the satellite were established; secondly, the EKF filter algorithm and the observability analysis method of the piecewise constant system (PWCS) were discussed; finally, through a comprehensive analysis of the XPNAV-1 satellite observation data, pulsar coverage of satellite orbits and observability of system status, the autonomous orbit determination algorithm experiment was carried out. The test results show that the filtering process of the autonomous orbit determination algorithm based on EKF converges, which verifies the rationality and effectiveness of the algorithm.

Key words: XPNAV-1 satellite, autonomous navigation, dynamic orbit determination, nonlinear filter, observability analysis