Target tracking analysis with Kalman filter in case of model mismatch

There is a mismatch between the estimator and the target models when the target movements include maneuvers. This leads to very high estimation errors and reduces the performance of tracker. In order to solve this problem, there are mainly two different approaches. In this study, to investigate the success of these approaches, different maneuvering movements of a single target were produced by means of simulation. For these maneuvers, the performance of the modified Kalman filters with the added process noise adaptation and the interacting multiple model filter has been examined. The results are presented comparatively.