Leibniz Universität Hannover Main Leibniz University Website
Deutsch Contact Deutsch
Collaborative Research Centre 1463
 
Collaborative Research Center 1463 - Integrated Design and Operation Methodology for Offshore Megastructures Research Project area C
Subproject C2

C2 – System identification and artificial intelligence in data-based structural health monitoring

Short title: System identification

The objective of subproject C02 is to develop methods that enable the reliable detection of structural changes in offshore megastructures during operation in order to minimize costly maintenance work. 

© http://www.sonnewindwaerme.de/

Structural changes can be, for example, undesired system changes (ice on the rotor blade) or potential damage (cracks). Here, the use of physically based finite element models is deliberately avoided. Instead, output only methods of system identification are used, for which only the measured response variables (accelerations, strains) and measurements of the environmental and operational conditions (temperature, rotor speed, wind speed, etc.) are considered.

The main challenges are that damage, or other critical system changes, cannot be measured directly, but can only be determined or estimated indirectly using damage-sensitive features. However, for a complex system such as an offshore wind turbine (OWEA), the number of feasible measurement points is small compared to the number of degrees of freedom of the structure. Thus, there is an information deficit. In addition, the structural dynamic behavior strongly depends on the environmental and operational conditions (EOCs). For example, interactions with aerodynamic forces occur at the rotor, which can lead to nonlinear vibration states. Therefore, classical methods of system identification, which assume linearity and time invariance of the metrologically observed system, cannot be readily applied. Moreover, a system change due to changing EOCs can easily be mistaken for damage.

On the other hand, in practical structural health monitoring (SHM), methods are urgently needed that allow rapid intervention in case of possible damage, but avoid unnecessary maintenance assignments due to false detections. Such maintenance operations are to be minimized as far as possible, especially in the offshore sector, due to the high accruing costs. The working hypothesis of the subproject (SP) is that a suitable combination of operational modal analysis (OMA) methods with physically interpretable state parameters and mathematical models from artificial neural networks (such as in Figure 1) can provide a suitable basis for structural monitoring of an offshore megastructure during operation.     
Fig. 1: Nonlinear autoregressive neural network (NARN)
Fig. 2: Modular SHM concept

The main innovation of the SP is the combination of robust algorithms based on white and black box models in a multivariate grey box SHM approach. Figure 2 shows a possible structure of a SHM- framework.    

This concept will be tested and validated gradually on databases of structures of different complexity. These structures range from a 9m high lattice tower standing on an open field, which can be reversibly damaged (see Figure 3), up to a real 3.4MW onshore wind turbine with a rotor diameter of 104m and a hub height of 125m (see Figure 4).    

© ISD/ LUH
Fig. 3: SHM Test structure – the lattice tower "LUMO"
© ISD/ LUH
Fig. 4: Onshore wind turbine equipped with sensors for research purposes

Publications

Subproject Management

Prof. Dr.-Ing. habil. Raimund Rolfes
Phone
+49 511 762 2992
Fax
+49 511 762 2236
Email
r.rolfesisd.uni-hannover.de
Web
http://www.isd.uni-hannover.de
Address
Appelstraße 9a
30167 Hannover
Building
3408
Room
426
Phone
+49 511 762 2992
Fax
+49 511 762 2236
Email
r.rolfesisd.uni-hannover.de
Web
http://www.isd.uni-hannover.de
Address
Appelstraße 9a
30167 Hannover
Building
3408
Room
426

Staff

Dr.-Ing. Tanja Grießmann
Phone
+49 511 762 2247
Fax
+49 511 762 2236
Email
t.griessmannisd.uni-hannover.de
Address
Appelstraße 9a
30167 Hannover
Building
3408
Room
429
Phone
+49 511 762 2247
Fax
+49 511 762 2236
Email
t.griessmannisd.uni-hannover.de
Address
Appelstraße 9a
30167 Hannover
Building
3408
Room
429
Sören Kai Möller
Phone
+49 511 762 3083
Fax
+49 511 762 2236
Email
s.moellerisd.uni-hannover.de
Address
Appelstraße 9a
30167 Hannover
Building
3408
Room
418
Phone
+49 511 762 3083
Fax
+49 511 762 2236
Email
s.moellerisd.uni-hannover.de
Address
Appelstraße 9a
30167 Hannover
Building
3408
Room
418

Last Change: 07.01.25 Print

Member Institutions
Contact us
Contact