Fault Detection and Identification for WindturbineSystems: a closed-loop analysis

During my Master's thesis project, I investigated model-based Fault Detection and Identification (FDI) methods for application in (offshore) wind turbines. The project was issued by NOVEM in close cooperation with ECN, and involved the modeling of a HAWT windturbine, a literature study of the state-of-the-art of FDI methods that have been applied up till now, and a study of system identification and observer-based FDI-methods. The latter used the Kalman filter and the IMM estimator.

Since offshore wind energy is a booming business, the field of FDI is very interesting. Maintenance costs generally increase due to the remote location, so a reliable FDI scheme can save a lot of money. Also, the availability of the wind turbine can be increased: monitoring the condition of the turbine allows that a component deterioration is detected in a very early stage, so that the component can be replaced before a major failure occurs.

On June 20, 2002, I obtained my M.Sc. degree in Applied Physics. My final thesis is available online. The m-files are available as well, but the latter are only accessible to the graduation committee and participants in the NOVEM project; If you fit into this category, please contact the author for a username/password combination.

Stijn Donders
Update: July 24, 2002