Dissertation Defense

Aaron Radke — Fri, 08 Dec 2006 19:12:02 +0000

Great news! the dissertation defense on December 7th, 2006, 10am was a success! It was a long time coming. The current committee version of the dissertation and the beautiful Keynote slides from the presentation are available for download.

Thank you to everyone who has helped me accomplish this with much encouragement.

I want to give special thanks to Dr. Gao for much grace to help me grow. To have him as a teacher, guide and friend is a privilege few get to experience. I am also thankful for the peer advice from my lab partners, especially Rob Miklosovic and Brian Fast. Many others, including Frank Goforth, Qing Zeng, Wankun Zhou, Jeff Csank and Gang Tian were always there to review, challenge and support ideas.

Aside from technical help, this project and my degree was impossible without the continued support and encouragement from my family. Above all and especially these years, I thank God for the increasingly revealed grace given to me through Jesus. There is little I deserve, yet I have been given an infinite treasure to have joy in being considered one of His children.

The research is stimulating in its own right, however the process and non technical items that I learned along the way are more significant. Discipline, planning and procrastination need to be taken up with a fight. Reading, writing and arithmetic are fundamental basics that I am just beginning to learn.

Here is the abstract:

This research focuses on the disturbance estimation problem. A survey of existing techniques is first conducted, where extensive plant knowledge for state estimation is contrasted with less plant knowledge for disturbance estimation. To advance the latter methodology and to reduce over-dependence on model information in the former, the boundary between a model and disturbance is redrawn. With this redefined boundary, existing methods are improved. Furthermore, to make the estimators optimal as well as practical, a new concept of practical optimality is introduced. Several novel design techniques are developed and initial results show advantages over current disturbance estimators. Finally, the disturbance estimation techniques are applied to health monitoring of dynamic systems. Two engineering examples are used to demonstrate the effectiveness and practicality of the new methods, including the health monitoring of a servomechanism and the fault diagnosis of a three-tank level control system.

Life's End

Dissertation Defense

Dissertation Defense