Robust SCKF Filtering Method for MINS/GPS In-Motion Alignment.

Area of Science:

• Control Systems Engineering
• Aerospace Engineering
• Signal Processing

Background:

• Accurate state estimation is critical for guided weapon systems, especially under dynamic flight conditions.
• Traditional Kalman filters struggle with rapidly changing system noise, leading to reduced estimation accuracy.
• Adaptive filtering techniques are needed to handle uncertainties and improve robustness in real-time applications.

Purpose of the Study:

• To introduce a novel Multiple Strong Tracking Adaptive Square-Root Cubature Kalman Filter (MSTASCKF).
• To enhance real-time system noise estimation and adapt the system noise variance matrix (Q).
• To improve the stability and accuracy of attitude estimation in the presence of random vibrations.

Main Methods:

• Developed an MSTASCKF based on the Sage-Husa filter framework.
• Incorporated a multi-fading factor to dynamically adjust the Q value and posterior error variance matrix.
• Utilized residual error orthogonality for filtering stability.
• Conducted semi-physical simulations using a vibration test with a shaking table to mimic flight conditions.

Main Results:

• The proposed MSTASCKF effectively estimates system noise in real-time.
• The adaptive adjustment of the Q matrix maintained filtering accuracy despite changing noise.
• The multi-fading factor ensured residual error orthogonality and filtering stability.
• Simulation results demonstrated a significant reduction in attitude estimation error caused by random vibration.

Conclusions:

• The MSTASCKF offers superior performance in handling time-varying noise compared to conventional methods.
• This adaptive filtering approach enhances the robustness and accuracy of attitude estimation for guided weapons.
• The filter's ability to adapt to dynamic noise is vital for maintaining guidance precision in challenging environments.