Bayesian Estimation for Tracking of Spiraling Reentry Vehicles

Abstract

A physics-based dynamics model of a noncooperative axisymmetric spiraling endoatmospheric reentry vehicle is presented for use in one of several model-based Bayesian state estimation frameworks, namely the extended Kalman filter and the particle filter. An analysis of the trajectory characteristics using elements from differential geometry lead to a relationship of the state of the vehicle to the spiraling motion. Simulated Earth entry trajectories illustrate the spiraling behavior that can be described by the model. The extended Kalman filter and the particle filter frameworks are employed to estimate the states and spiraling characteristics in the presence of model and environmental uncertainties and measurement errors. We assume ground based radar(s) provide range, range rate, azimuth and elevation measurements. Error budgets and sensitivity plots are provided depicted the sources of uncertainty with greatest impact on the ability to estimate the position and spiraling frequency. It is shown that the extended Kalman filter and the particle filter have comparable state estimation performance and that the largest contributor to state estimation errors stem from uncertainty in the lift angle and rate of the spiraling vehicle.