Study on the Robust Filter Method of SINS/DVL Integrated Navigation Systems in a Complex Underwater Environment

Area of Science:

• Navigation Systems Engineering
• Signal Processing
• Control Theory

Background:

• Strapdown Inertial Navigation System (SINS)/Doppler Velocity Log (DVL) integrated systems face navigation accuracy degradation due to measurement noise characteristics.
• Traditional adaptive filtering algorithms often encounter non-positive definite matrix issues, hindering accurate noise estimation.
• Accurate estimation of measurement noise is critical for robust performance in integrated navigation.

Purpose of the Study:

• To develop an improved adaptive filtering algorithm for SINS/DVL systems to enhance navigation accuracy and stability.
• To address the challenge of measurement noise impacting integrated navigation performance.
• To overcome limitations of existing adaptive filtering methods, particularly the non-positive definite matrix problem.

Main Methods:

• An improved adaptive filtering algorithm based on the Sage-Husa adaptive Kalman filtering algorithm is proposed.
• The method incorporates upper and lower thresholds, determined by a discrimination factor, to manage abnormal measurement data.
• Innovation covariance is adjusted, and measurement noise is re-estimated using a decision factor based on the innovation.

Main Results:

• The proposed algorithm demonstrates superior navigation accuracy and stability compared to the classical Kalman filter (KF).
• Filtering performance significantly surpasses that of the Sage-Husa algorithm.
• Simulation results show a 49.44% reduction in relative position error compared to the Sage-Husa filtering method.

Conclusions:

• The improved adaptive filtering algorithm effectively enhances navigation accuracy and stability in SINS/DVL systems.
• The proposed method provides a robust solution for dealing with measurement noise and abnormal data.
• This algorithm offers a significant advancement over existing adaptive filtering techniques for integrated navigation.