IMM algorithm based on the analytic solution of steady state Kalman filter for radar target tracking

Recently, the interacting multiple model (IMM) algorithm based on the steady state Kalman filters has been proposed as a very attractive method for real-time implementation. But when the tracking filter is designed in the Cartesian coordinates, the covariance matrix of radar measurement error varies according to the range and bearing of the target. Therefore, the steady state Kalman gain and the covariance matrix calculated off-line may become inappropriate. In this paper, the IMM tracker is formulated in the Cartesian coordinate frame based on the analytic solution of the steady state Kalman filter in which gain and covariance matrix are calculated on-line. The performance of the proposed approach is compared with the conventional IMM tracker in terms of the root mean square error (RMSE) and the normalized position error (NPE) via simulation. The simulation results indicate that this approach not only improves the accuracy but also reduces computational load.