Tracking Highly Maneuvering Targets Using Fuzzy Covariance Matrix Presetting

In tins paper a new covariance presetting scheme is presented to overcome some drawbacks of highly maneuvering target tracking problems by using a fuzzy logic method Tins method IS based on fuzzy covariance matrix presetting The scheme includes an estimation part that uses a modified Kalman filter combined with a fuzzy logic part to improve the tracking performance of highly maneuvering targets. The fuzzy part, IS designed based on deviation of the last, target direction. The results are compared with two conventional covariance matrix presetting methods with and without jerk model The proposed method is rather robust with respect to the values of sampling time Simulation results show a superior performance of the proposed covariance presetting method in various scenarios.