Electrical and Computer Engineering ETDs

Publication Date

2-7-2011

Abstract

A common problem in developing new gimbal products is predicting performance. At the beginning of the design stage, typically the proposal writing stage, it is critical to be able to anticipate the performance of a design, which may still be very roughly defined. Static performance metrics such as pointing accuracy are easier to predict and can be related to position sensor resolution and compliance in the structure and drive. Dynamic performance metrics such as rate tracking and bandwidth are much more difficult to estimate. These dynamic parameters will serve as the results of this paper. A gimbal model is usually developed in this early stage of design. A model is built for each axis as cross-coupling is not usually significant. The development of each model is inherently dependent on the integrity of the parameters used. Some parameters may be easily accessible and clearly defined, such as manufacturer specifications for commercial-off-the-shelf (COTS) components. An example of these would be motors. A motor datasheet will usually include specifications such as winding resistance, torque constant, etc. Some parameters have to estimated such as drive friction, structural rigidity and drive parameters. There is also much less certainty in these estimates due to their dependence on the integrated, final system. Building an accurate and sufficient model of the system is a challenging task. This thesis developments and validates a single axis gimbal model. Following this is a recursion on the model parameters based upon empirical data. Finally, application of different control laws are evaluated in simulation on the model. First, a classical output feedback law is implemented. Secondly, a state observer is implemented with state feedback. State feedback coefficients are found using both pole placement (PP) algorithms and linear quadratic regulation (LQR) optimal control formulas.

Keywords

Linear control systems--Mathematical models, Feedback control systems--Mathematical models, PID controllers--Mathematical models.

Sponsors

RIEtech Global, LLC

Document Type

Thesis

Language

English

Degree Name

Electrical Engineering

Level of Degree

Masters

Department Name

Electrical and Computer Engineering

First Advisor

Abdallah, Chaouki

First Committee Member (Chair)

Jayaweera, Sudharman

Second Committee Member

Fierro, Rafael

Share

COinS