Improved trajectory estimation of maneuvering reentry vehicles using nonlinear filter based on the Pontryagin minimum principle

This paper introduces a new nonlinear filter that is used for the real-time estimation of the trajectory of a maneuvering reentry vehicle (MARV) from its radar observations. The performance, as measured by the point error, is compared to that of the conventional extended Kalman filter (EKF). The proposed nonlinear filter is based on optimal control concepts, specifically, the Pontryagin minimum principle. Using these concepts, the unknown maneuvering forces are treated as controllers that drive the MARV dynamics to follow or track the noisy observed path. This treatment is different from the approach used with EKF, where the unknown forces are considered as Wiener processes and new states are augmented to the MARV states. The computational time for the proposed nonlinear filter, for the cases studied, is about 20 percent that of EKF, which is a substational improvement. The relationship between EKF and the proposed nonlinear filter is also discussed.