

Automated System Identification and Validation of Numerical Models of Offshore Wind Turbines as Basis for SHM-Analysis
Abstract
System identification by measured data is an important procedure, necessary for many applications within the field of structural health monitoring. The resulting modal properties can be used for model- validation and updating, load estimation and damage detection to name a few possibilities. An automated algorithm for identifying a structure’s modal properties from dynamic response data (accelerations) using Auto- Regressive (AR) models is presented here. New parameters to estimate the ARmodel’s order a priori and to extract the physical from the mathematical models are introduced. Furthermore, a tool for automated model validation using several iterative techniques is presented as well. The algorithms are applied to data from two offshore structures. System identification and model updating are accessible through a graphic user interface (GUI) based on MATLAB.