Abstract: An attitude determination method for a deep-space mission spacecraft is derived and its performance is analyzed.The attitude determination system is composed of inertial measurement units(IMU),an X-ray Pulsar scanner,onboard computers and electronic devices.IMU data is processed to maintain real-time knowledge of spacecraft attitude relative to an inertial reference frame.X-ray pulsar scanner data is processed using unscented Kalman filter(UKF) to estimate and correct the attitude determination errors and the gyro drift compensation errors.The results of two X-ray pulsars availability analysis for deep-space attitude determination mission are presented.Linear covariance analysis techniques are used to evaluate nominal attitude determination performance,the effects of sensor measurement accuracy variations,and the effects of gyros misalignment errors.A semi-physical simulation scheme is devised to verify the integrated attitude determination algorithm based on the data from NASA HEASARC.Results of a nonlinear simulation analysis of attitude determination performance are also shown.Compared to only using pulsars or IMU,the attitude determination accuracy of integrated algorithm is higher,and the performance of processing nonlinear state functions for UKF is superior to EKF.It can satisfy the system requirements of high precision and stabilization.