Maximum correntropy square-root cubature Kalman filter with application to SINS/GPS integrated systems

Area of Science:

• Engineering
• Control Systems
• Signal Processing

Background:

• Cubature Kalman Filter (CKF) and its square-root version (SCKF) are effective for state estimation in nonlinear systems under Gaussian noise.
• Traditional filters struggle with non-Gaussian noises, especially heavy-tailed impulsive disturbances, leading to performance degradation.

Purpose of the Study:

• To develop a novel nonlinear filter that enhances robustness against heavy-tailed non-Gaussian noises.
• To improve state estimation accuracy in challenging noise conditions by incorporating the Maximum Correntropy Criterion (MCC).

Main Methods:

• Proposed the Maximum Correntropy Square-Root Cubature Kalman Filter (MCSCKF).
• Utilized the Maximum Correntropy Criterion (MCC) to replace the Minimum Mean Square Error (MMSE) criterion for improved robustness.
• Introduced a judgment condition to prevent numerical issues.

Main Results:

• The MCSCKF retains the advantages of the SCKF while demonstrating superior robust performance in non-Gaussian noise scenarios.
• Evaluated performance using two examples, including SINS/GPS integrated systems, confirming the filter's effectiveness.
• The proposed filter shows desirable performance against heavy-tailed non-Gaussian noises.

Conclusions:

• The MCSCKF offers a robust solution for state estimation in nonlinear systems contaminated by heavy-tailed impulsive noises.
• The filter provides a valuable alternative to traditional methods when dealing with non-Gaussian noise environments.
• The study validates the practical applicability and effectiveness of the MCSCKF in complex systems like SINS/GPS.