Three-Dimensional Bearings-Only Target Tracking: Comparison of Few Sigma Point Kalman Filters

Three-dimensional bearings-only target tracking problem is considered in this work. To solve this problem, highly accurate nonlinear filtering algorithms are required. This is because the nonlinearities in the noise corrupted measurements, combined with the other uncertainties, make the estimation problem highly challenging. In this work, the extended Kalman filter (EKF), the cubature Kalman filter (CKF), the unscented Kalman filter (UKF), and the new sigma point Kalman filter (NSKF) are used to solve this state estimation problem. The performance criteria chosen are root mean square error (RMSE) in position and velocity. To have a more meaningful study, RMSE at the end of the observation period and time-averaged RMSE after the observer manoeuvre is calculated, for varying measurement noise and initial uncertainty.