2D and 3D Angles-Only Target Tracking Based on Maximum Correntropy Kalman Filters

Area of Science:

• Signal Processing
• Estimation Theory
• Robotics

Background:

• Angles-only target tracking (AoT) is crucial for navigation and surveillance.
• Conventional estimators struggle with non-Gaussian noise and outliers.
• Impulsive noise degrades the performance of minimum mean square error (MMSE) estimators.

Purpose of the Study:

• To develop robust estimation algorithms for AoT in non-Gaussian environments.
• To address the limitations of MMSE estimators in the presence of outliers.
• To introduce a maximum correntropy criterion (MCC) based framework for improved AoT accuracy.

Main Methods:

• Developed three novel estimation algorithms: MC-UKF-CK, MC-NSKF-GK, and MC-NSKF-CK.
• Utilized sigma point filters, including the unscented Kalman filter (UKF).
• Employed Gaussian and Cauchy kernels within the MCC framework.

Main Results:

• The proposed algorithms demonstrated superior estimation accuracy compared to conventional methods.
• Evaluated performance using root-mean-square error (RMSE) in position and % track loss.
• Simulations in 2D and 3D AoT scenarios confirmed improved robustness against non-Gaussian noise.

Conclusions:

• The MCC-based framework provides a robust solution for AoT problems with non-Gaussian measurement noise.
• The developed algorithms (MC-UKF-CK, MC-NSKF-GK, MC-NSKF-CK) offer enhanced accuracy and reliability.
• This research advances AoT capabilities in challenging, real-world environments.