Date of Award
Doctor of Philosophy (PhD)
Mechanical & Aerospace Engineering
Accurate state information is crucial for control of flexible space structures in which the state feedback strategy is used. The performance of a state estimator relies on accurate knowledge about both the system and its disturbances, which are represented by system model and noise covariances respectively. For flexible space structures, due to their great flexibility, obtaining good models from ground testing is not possible. In addition, the characteristics of the systems in operation may vary due to temperature gradient, reorientation, and deterioration of material, etc. Moreover, the disturbances during operation are usually not known. Therefore, adaptive methods for system identification and state estimation are desirable for control of flexible space structures. This dissertation solves the state estimation problem under three situations: having system model and noise covariances, having system model but no noise covariances, having neither system model nor noise covariances. Recursive least-squares techniques, which require no initial knowledge of the system and noises, are used to identify a matrix polynomial model of the system, then a state space model and the corresponding optimal steady state Kalman filter gain are calculated from the coefficients of the identified matrix polynomial model. The derived methods are suitable for on-board adaptive applications. Experimental example is included to validate the derivations.
"Integrated System Identification and Adaptive State Estimation for Control of Flexible Space Structures"
(1991). Doctor of Philosophy (PhD), dissertation, Mechanical & Aerospace Engineering, Old Dominion University, DOI: 10.25777/nxry-sj86