Acta Electronica Sinica (selected articles)

The theory and simulation results of employing a Kalman filter in an airborne fire control radar tracking system was studied. Problems such as target dynamics modeling, linear filtering and simulation, linear filtering approximation, sensitivity simulation, maneuvering target tracking and adaptivity were considered. A modified Kalman filter for tracking a maneuvering target was presented. It detects a target maneuver by judging whether there is a bias in the observation residue. This estimate was used to correct the state prediction and error covariance. It worked as a single Kalman filter with its maneuver acceleration command at zero. A better compromise between the steady state filtering accuracy and fast response to maneuver was reached. Computer simulation results show that the accuracy of the filter is slightly superior to that of a complicated filter bank.