Joshua T. Horwood (Numerica Corporation), Nathan D. Aragon (Numerica Corporation), Aubrey B. Poore (Numerica Corporation)
Keywords: Covariance Consistency, Uncertainty Consistency, Nonlinear Filtering, Unscented Kalman Filter, Gaussian Mixture,
Abstract:
The accurate and consistent representation of a space object’s uncertainty is essential in the problems of data association (correlation), conjunction analysis, sensor resource management, and anomaly detection. While standard Kalman-based filtering algorithms, Gaussian assumptions, and covariance-weighted metrics are very effective in data-rich tracking environments, their use in the data sparse environment of space surveillance is largely inadequate. It is shown how improved uncertainty consistency can be maintained using the higher fidelity Edgeworth or adaptive Gaussian mixture filters in an orbital element space and how statistics beyond a Gaussian state and covariance can be represented correctly. A simulation scenario which considers the implications of correct uncertainty management to data association (correlation) and anomaly detection is presented.
Date of Conference: September 14-17, 2010
Track: Astrodynamics