Nonlinear Uncertainty Propagation of Satellite State Error for Tracking and Conjunction Risk Assessment
Abstract
Developed a new approach to uncertainty quantification in non-linear systems with application to orbit trajectory prediction and satellite conjunction analysis. Statistical approach utilizes novel methods to build better uncertainty state characterization in the context of rare event prediction while keeping computational expense low. Previous models arent suited to predict low probability regions that are of importance to accurately calculating conjunction risk of satellites. Prior models utilize significant but sometimes arbitrary buffers to account for the unknown true statistical distribution of satellite position, which manifests as a result of the non-linear properties of satellite motion. The method we developed propagates uncertainty contours, and utilizes analysis of the non-linear system to better sample the uncertainty volume, and optimizes the characterization of tail probabilities. This method can be used to assess the risk of rare events, namely satellite conjunction. During this one year project we developed the basic framework for propagating error distributions and quantifying the resultant approximation error in the context of dynamic equations describing orbital trajectories.
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 18, 2017
- Accession Number
- AD1046932
Entities
People
- Alfred Hero
- Audelia Wittbrodt
Organizations
- University of Michigan