Journal of Shanghai Jiao Tong University(Science) ›› 2020, Vol. 25 ›› Issue (4): 434-440.doi: 10.1007/s12204-020-2208-x

• • 上一篇    下一篇

Adaptive Dual Wavelet Threshold Denoising Function Combined with Allan Variance for Tuning FOG-SINS Filter

BESSAAD Nassim, BAO Qilian (鲍其莲), SUN Shuodong (孙朔冬), DU Yuding (杜雨丁), LIU Lin (刘林), HASSAN Mahmood Ul   

  1.  (1. School of Electronic Information and Electrical Engineering, Shanghai Jiao Tong University, Shanghai 200240, China;
    2. Shanghai Aerospace Control Technology Institute, Shanghai 201109, China)
  • 出版日期:2020-08-28 发布日期:2020-07-29
  • 通讯作者: BAO Qilian (鲍其莲) E-mail:qlbao@sjtu.edu.cn

Adaptive Dual Wavelet Threshold Denoising Function Combined with Allan Variance for Tuning FOG-SINS Filter

BESSAAD Nassim, BAO Qilian (鲍其莲), SUN Shuodong (孙朔冬), DU Yuding (杜雨丁), LIU Lin (刘林), HASSAN Mahmood Ul   

  1. (1. School of Electronic Information and Electrical Engineering, Shanghai Jiao Tong University, Shanghai 200240, China;
    2. Shanghai Aerospace Control Technology Institute, Shanghai 201109, China)
  • Online:2020-08-28 Published:2020-07-29
  • Contact: BAO Qilian (鲍其莲) E-mail:qlbao@sjtu.edu.cn

摘要:  Allan variance (AV) stochastic process identification method for inertial sensors has successfully combined
the wavelet transform denoising scheme. However, the latter usually employs a traditional hard threshold
or soft threshold that presents some mathematical problems. An adaptive dual threshold for discrete wavelet
transform (DWT) denoising function overcomes the disadvantages of traditional approaches. Assume that two
thresholds for noise and signal and special fuzzy evaluation function for the signal with range between the two
thresholds assure continuity and overcome previous difficulties. On the basis of AV, an application for strap-down
inertial navigation system (SINS) stochastic model extraction assures more efficient tuning of the augmented 21-
state improved exact modeling Kalman filter (IEMKF) states. The experimental results show that the proposed
algorithm is superior in denoising performance. Furthermore, the improved filter estimation of navigation solution
is better than that of conventional Kalman filter (CKF).

关键词: Allan variance (AV), discrete wavelet transform (DWT), adaptive dual threshold, fiber optic gyroscope
(FOG),
strap-down inertial navigation system (SINS), exact modeling filter

Abstract:  Allan variance (AV) stochastic process identification method for inertial sensors has successfully combined
the wavelet transform denoising scheme. However, the latter usually employs a traditional hard threshold
or soft threshold that presents some mathematical problems. An adaptive dual threshold for discrete wavelet
transform (DWT) denoising function overcomes the disadvantages of traditional approaches. Assume that two
thresholds for noise and signal and special fuzzy evaluation function for the signal with range between the two
thresholds assure continuity and overcome previous difficulties. On the basis of AV, an application for strap-down
inertial navigation system (SINS) stochastic model extraction assures more efficient tuning of the augmented 21-
state improved exact modeling Kalman filter (IEMKF) states. The experimental results show that the proposed
algorithm is superior in denoising performance. Furthermore, the improved filter estimation of navigation solution
is better than that of conventional Kalman filter (CKF).

Key words: Allan variance (AV), discrete wavelet transform (DWT), adaptive dual threshold, fiber optic gyroscope
(FOG),
strap-down inertial navigation system (SINS), exact modeling filter

中图分类号: