Particle Filter-based Pure Bearing Tracking Algorithm

In this paper, we investigate a particle filter-based pure bearing tracking algorithm and conduct comparative analysis with traditional Kalman Filter (KF) and Extended Kalman Filter (EKF) methods. The implementation utilizes sequential Monte Carlo methods where particles represent potential target states, with importance sampling and resampling techniques handling the nonlinear bearing-only measurements. Experimental results validate the algorithm's effectiveness, particularly in complex environments with non-Gaussian noise and multimodal distributions. The study further examines the algorithm's advantages in handling nonlinear systems and its computational limitations, while proposing enhancement directions including adaptive particle number adjustment and optimized proposal distributions. Overall, this research provides novel approaches for achieving high-precision target tracking through sophisticated particle filtering techniques.