Adaptive Estimation and Heuristic Optimization of Nonlinear Spacecraft Attitude Dynamics
Abstract
For spacecraft conducting on-orbit operations, changes to the structure of the spacecraft are not uncommon. These planned or unanticipated changes in inertia properties couple with the spacecraft's attitude dynamics and typically require estimation. For systems with time-varying inertia parameters, multiple model adaptive estimation (MMAE) routines can be utilized for parameter and state estimates. MMAE algorithms involve constructing a bank of recursive estimators, each assuming a different hypothesis for the systems dynamics. This research has three distinct, but related, contributions to satellite attitude dynamics and estimation. In the first part of this research, MMAE routines employing parallel banks of unscented attitude filters are applied to analytical models of spacecraft with time-varying mass moments of inertia (MOI), with the objective of estimating the MOI and classifying the spacecraft's behavior. New adaptive estimation techniques were either modified or developed that can detect discontinuities in MOI up to 98 % of the time in the specific problem scenario. Second, heuristic optimization techniques and numerical methods are applied to Wahba's single-frame attitude estimation problem, decreasing computation time by an average of nearly 67%. Finally, this research poses MOI estimation as an ODE parameter identification problem, achieving successful numerical estimates through shooting methods and exploiting the polhodes of rigid body motion with results, on average, to be within < 1 % to 5 % of the true MOI values.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 15, 2016
- Accession Number
- AD1018254
Entities
People
- Joshuah A. Hess
Organizations
- Air Force Institute of Technology