Application of Adaptive Filtering to an Attitude Determination Problem.
Abstract
A gyro is mounted on a satellite to measure body angular rates between star sightings. A Kalman Filter is used to update the estimate of attitude which is degraded between star sightings because of random gyro drift. The estimated autocorrelation of the measurement residuals are employed to do the following: (1) Detect changes in the noise parameters of either the gyro or the star sensor, (2) Estimate the true estimate covariance, and (3) Estimate the optimal Kalman gain or adapt the filter to the new noise parameters. Numerical examples are included. It is important to emphasize the following: (1) The filter performance can be estimated by observing only the measurement residuals, and (2) The filter parameters can be adapted to improve the performance, if desired. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 12, 1970
- Accession Number
- AD0751285
Entities
People
- Philip L. Smith
Organizations
- The Aerospace Corporation