A decoupled adaptive tracking filter for real applications

A decoupled adaptive tracking filter for maneuvering targets under nonlinear measurement environment is presented. A modified input estimation method is implemented in a line-of-sight Cartesian coordinate system. Since the proposed filter has a decoupled structure, computation is drastically reduced without degrading performance. Also shown is that, due to small cross-correlation between pseudomeasurement noises, accuracy of maneuver detection is improved.