Variational Bayesian Based Adaptive Shifted Rayleigh Filter for Bearings-Only Tracking in Clutters

Area of Science:

• Signal Processing
• Estimation Theory
• Target Tracking

Background:

• Bearings-only target tracking is crucial in various applications but challenging in cluttered environments.
• Traditional filters like the shifted Rayleigh filter (SRF) assume known clutter probability, limiting performance when this is uncertain.
• Uncertainty in clutter probability significantly degrades the performance of conventional tracking algorithms.

Purpose of the Study:

• To develop an improved bearings-only target tracking filter that addresses uncertain clutter probability.
• To enhance tracking accuracy, continuity, and robustness in complex and adverse scenarios.
• To provide a robust solution for scenarios where clutter characteristics are not precisely known.

Main Methods:

• A variational Bayesian approach is employed to develop an adaptive shifted Rayleigh filter (VB-SRF).
• The VB-SRF jointly estimates the target state and the clutter probability, accommodating uncertainty.
• Performance is evaluated through simulations comparing VB-SRF against SRF and Probability Data Association (PDA)-based filters.

Main Results:

• The proposed VB-SRF demonstrates superior performance compared to traditional SRF and PDA-based filters.
• Significant improvements are observed in track continuity, accuracy, and robustness, particularly in complex adverse scenarios.
• The VB-SRF achieves enhanced performance with a slightly increased computational complexity.

Conclusions:

• The variational Bayesian-based adaptive shifted Rayleigh filter (VB-SRF) offers a robust solution for bearings-only target tracking in uncertain clutter.
• VB-SRF effectively handles uncertainty in clutter probability, outperforming existing methods in challenging conditions.
• This adaptive approach provides a valuable advancement for practical target tracking systems facing environmental uncertainties.