导航

ACTA AERONAUTICAET ASTRONAUTICA SINICA ›› 2019, Vol. 40 ›› Issue (7): 322884-322884.doi: 10.7527/S1000-6893.2019.22884

• Electronics and Electrical Engineering and Control • Previous Articles     Next Articles

Relative navigation algorithm for non-cooperative target with adaptive modification of multiplicative noise

ZHU Yunfeng, SUN Yongrong, ZHAO Wei, HUANG Bin, WU Ling   

  1. College of Automation Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing 210016, China
  • Received:2019-01-02 Revised:2019-02-14 Online:2019-07-15 Published:2019-02-22
  • Supported by:
    National Natural Science Foundation of China(61374115);Fundamental Research Funds for the Central Universities(NZ2016104);the Jiangsu Province Policy Guidance Class Funds for Production Study-Research Cooperation (BY206003-16)

Abstract: The task of Unmanned Aerial Vehicle (UAV) situational awareness is to use airborne sensors to identify and guide targets in unknown environments. To solve the relative navigation problem between UAV and non-cooperative target at medium and long distances, a relative state estimation algorithm based on angle and distance measurements is proposed. On the basis of existing filtering algorithms, in order to improve the accuracy and stability, this paper improves the Iterative Extended Kalman Filter (IEKF) algorithm by using the Levenberg-Marquardt (LM) optimization. This paper proposes a Levenberg-Marquardt Iterative Extended Kalman Filter (LM-IEKF) algorithm, deducing the state updating equation and the recursive formula of covariance matrix in the iteration process of the LM-IEKF algorithm. The multiplicative noise introduced by the distance sensor due to the signal correlation characteristics makes the existing additive noise model difficult to adapt. Therefore, a Modified LM-IEKF method based on adaptive correction of measurement noise statistics is proposed. The filtering accuracy is improved by updating the noise matrix online in real time. Meanwhile, the results of fading memory index smoothing estimation are set up. Verification results of the algorithm show that compared with the existing EKF (Extended Kalman Filter) and IEKF algorithms, the LM-IEKF algorithm has better performance in additive noise only situation. In the case of containing multiplicative noise, Modified LM-IEKF can effectively estimate the measurement noise. Compared with the widely used EKF algorithm, the accuracy of integrated relative position and relative velocity improved by 10% and 23%.

Key words: relative navigation, Iterative Extended Kalman Filter (IEKF), Levenberg-Marquardt optimization, multiplicative noise, non-cooperative target

CLC Number: