Stochastic Feedback Based Continuous-Discrete Cubature Kalman Filtering for Bearings-Only Tracking

Area of Science:

• Navigation and Guidance
• Estimation Theory
• Signal Processing

Background:

• Bearings-only tracking relies solely on angle measurements for target state estimation.
• Approximation errors in filtering processes (e.g., linearization) degrade tracking accuracy.
• Accurate state prediction is crucial for effective bearings-only target tracking.

Purpose of the Study:

• To propose an online covariance adaptation method for bearings-only tracking.
• To mitigate unpredictable approximation errors inherent in filtering.
• To enhance the accuracy and efficiency of target state estimation.

Main Methods:

• Developed an online covariance adaptation technique using posterior covariance information.
• Investigated the theoretical relationship between posterior and prior covariance.
• Integrated continuous-discrete cubature Kalman filtering with a covariance updating feedback rule.

Main Results:

• Demonstrated the effectiveness of posterior covariance for online adaptation.
• Successfully modified priori covariance online via a feedback rule.
• The proposed framework showed improved computational accuracy and efficiency.

Conclusions:

• Online covariance adaptation effectively reduces unpredictable errors in bearings-only tracking.
• The integrated framework enhances the performance of target tracking filters.
• The method offers a more accurate and computationally efficient solution for bearings-only navigation.