A nonlinear approach to the problem of tracking a maneuvering aircraft
is presented. A nonlinear aircraft maneuver model is proposed for sta
te estimation as well as prediction. The nonlinear model describes pla
nar trajectories. A geometric nonlinear filter (GNF) is presented. The
results show a significant overall improvement over the extended Kalm
an filter (EKF) approach. The GNF was found to be stable in most cases
where the EKF was not stable. For the stable cases, the tracking perf
ormance of the GNF compares favorably with the EKF. The GNF offers a s
ubstantial savings in computational time.