• Lasse Bonde Hansen
  • Michael Joseph Keenan Odena
4. semester, Sustainable Energy Engineering, Master (Master Programme)
This report details the development of diverse fault detection and diagnosis al- gorithms employed for detecting faults in cooling fans utilized for temperature management in control cabinets of wind turbines. To replicate the thermodynamic behavior of an actual cabinet, two mod- els were created: a single-zone model and a multi-zone model. These mod- els simulated and collected temperature measurements of both the cabinet and the nacelle. To identify the system para- meters, a calibration phase utilizing the Recursive Least Squares (RLS) method was implemented based on the temper- ature measurements. The acquired para- meters, along with the measurements, were then employed in the fault detection and diagnosis (FDD) algorithms, includ- ing a bank of Observers, Multiple Model Adaptive Estimation (MMAE), and Joint State estimation. Through these methods, reliable temperature estimation and pre- diction of the cooling fans’ health status within a certain range were achieved.
SpecialisationOffshore Energy Systems
Publication date1 Jun 2023
Number of pages69
External collaboratorSiemens
Frank Vandborg frank.vandborg@siemensgamesa.com
Information group
ID: 532431888