The robust estimation of mass properties for spacecraft in low-earth orbit /
The estimation algorithm to determine the mass ty Microfilm Inc. properties of a spacecraft in low-earth orbit using attitude and actuator measurements is investigated. The mass properties include the mass, the moments of inertia, and the center of mass location. A linear regression model, obtained...
| Main Author: | |
|---|---|
| Format: | Thesis Book |
| Language: | English |
| Published: |
[Place of publication not identified] :
[publisher not identified] ;
1998.
|
| Subjects: | |
| Online Access: | http://proxy.library.tamu.edu/login?url=http://proquest.umi.com/pqdweb?did=733039041&sid=1&Fmt=2&clientId=2945&RQT=309&VName=PQD |
| Summary: | The estimation algorithm to determine the mass ty Microfilm Inc. properties of a spacecraft in low-earth orbit using attitude and actuator measurements is investigated. The mass properties include the mass, the moments of inertia, and the center of mass location. A linear regression model, obtained from the rotational equations of motion and sensor dynamics, is formed to estimate the uncertain parameters. A parameter estimation algorithm solves for the unknown parameters in the regression model using the altered measurements. The regression model to solve for the moments of inertia is derived from the rotational equations of motion for a spacecraft, including the effect of the control moment gyros. The regression model for the center of mass location is derived from the inertial acceleration measured by the accelerometers. The mass cannot be solved from a linear regression model, but it can be determined, if a second spacecraft docks to the spacecraft, or a spacecraft component, like the robotic arm, moves to a new opinion. Measurement noise and modeling error in the regression model can cause a large parameter estimation error using traditional estimation algorithms. This error can be attenuated by profiteering the measurements using a smoothing algorithm to remove high frequency noise. The general recursive least squares algorithm is designed to be robust to measurement noise and modeling error by introducing a weighting parameter which will prevent a large correction to the previous parameter estimate, when the regression equation is numerically inconsistent due to these errors. The optimal weighting parameter is selected which will minimize the confidence interval width over a set of asymptotically unbiased parameter estimates. The mass properties estimation algorithm is tested using numerical simulations of an attitude maneuver for the International Space Station (ISS), under the influence of gravity-gradient, aerodynamic, and disturbance torques. A Lyapunov controller reorients the ISS to its torque equilibrium attitude. The moments of inertia can be estimated very well, when the measurements include Gaussian white noise. The estimation of the center of mass location from the accelerometer regression model is theoretically possible, but estimation errors from measurement noise and profiteering make it impractical for real spacecraft. |
|---|---|
| Item Description: | Vita. "Major Subject: Aerospace Engineering". |
| Physical Description: | xviii, 264 leaves : illustrations ; 28 cm. |
| Bibliography: | Includes bibliographical references (leaves 239-244). |