Adaptive filter based on Monte Carlo method to improve the non-linear target tracking in the radar system

Abstract

This article deals with the problem of degraded tracking performance of a high non-linear target in a radar system, well known by the divergence phenomenon. In our study, we aim to improve the target state estimation to imitate the tracking scenario as well as avoid the last cited undesirable phenomenon, generated during the non-linear measurements filtering, once using extended KALMAN filter. To overcome this issue, we have implemented a new approach based on the adaptive Monte Carlo (AMC) algorithm to replace the traditional method as is known by the extended KALMAN filter (EKF). The obtained experimental results showed a challenging remediation. Where, the AMC converges towards the accurate state estimation. Thus, more efficient than extended KALMAN filter. The experimental results prove that the designed system meets the objectives set for AMC referring to an experimental database obtained by a radar system, using MATLAB software development framework.