导航

ACTA AERONAUTICAET ASTRONAUTICA SINICA ›› 2016, Vol. 37 ›› Issue (4): 1305-1315.doi: 10.7527/S1000-6893.2015.0175

• Electronics and Control • Previous Articles     Next Articles

A robust extended Kalman filter algorithm for X-ray pulsar navigation system

LI Min, ZHANG Yingchun, GENG Yunhai, ZHU Baolong, LI Huayi   

  1. School of Astronautics, Harbin Institute of Technology, Harbin 150001, China
  • Received:2015-04-22 Revised:2015-06-07 Online:2016-04-15 Published:2015-07-02
  • Supported by:

    National Natural Science Foundation of China (61473297,61304237)

Abstract:

Considering that the traditional extended Kalman filter (EKF) algorithm has low accuracy and poor robustness in the application of the X-ray pulsar navigation system with uncertainty of model parameters, multiplicative noises and additive noises, a robust EKF algorithm is proposed in this paper. The equations for state prediction errors and state estimation errors are established. Then, state prediction error covariance and state estimation error covariance are obtained. However, as there are uncertainties in model parameters, the accurate value of state prediction error covariance and state estimation error covariance cannot be calculated. Hence, the upper bound of the state prediction error covariance and the upper bound of the state estimation error covariance are identified by using four matrix inequalities. The gain matrix is designed by minimizing the upper bound of the state estimation error covariance for all admissible uncertainties. The method is applied to X-ray pulsar navigation system, and simulation results show the effectiveness of the algorithm.

Key words: X-ray pulsar navigation system, uncertainty, nonlinear system filtering, robust extended Kalman filter, estimation accuracy

CLC Number: