Comparison of angle-only filtering algorithms in 3D using EKF, UKF, PF, PFF, and ensemble KF
Syamantak Datta Gupta, Jun Ye Yu, Mahendra K. Mallick, Mark Coates, Mark R. Morelande · International Conference on Information Fusion · 2015
In our previous work, we compared the performance of the extended Kalman filter (EKF), unscented Kalman filter (UKF), and particle filter (PF) for the angle-only filtering (AOF) problem in 3D using Cartesian coordinates and modified spherical coordinates (MSC) for the relative state vector. We found that the UKF-MSC and EKF-MSC had the best performance in accuracy, the UKF-MSC being slightly better than the EKF-MSC. The PF didn't perform well compared with the EKF and UKF and had a higher computational cost. In this work, we compare the performance of the particle flow filter (PFF) with the other filters for the AOF problem. In addition, we also analyze the performance of two versions of the ensemble Kalman filter (EnKF) in this comparative study. We present numerical results from Monte Carlo simulations to analyze the state estimation accuracy and computational cost of these filters.