State estimation of reentry vehicles from optical tracking data

A computer software package for generating filtered and smoothed estimates of the state of a reentry vehicle using optical tracking data is developed. Dynamic and observation models, a second-order filter algorithm with plant noise, a fixed-interval smoother algorithm for nonlinear systems, and an adaptive plant noise algorithm for solving divergence problems were derived. The operation of the software package is verified by generating estimates using three different data sets. The first case is based on a simulated trajectory of a reentry vehicle with a constant ballistic coefficient, the second case is based on a simulated trajectory of a vehicle whose ballistic coefficient is a parabolic function of altitude, and the third case is based on actual data with an unknown true trajectory. A Monte-Carlo study is conducted to determine the confidence in package performance when processing other data sets. Results are presented.