In many real-world applications changing environmental and operational conditions, e.g. temperature and loads, complicate the monitoring of a structure or make it impossible. Thus a need arises to compensate these influences in order to reduce the number of false alarms to a minimum. The knowledge of the environment enables us to incorporate these influences into a reliable statement about the health of the structure. Thus an improvement of damage detection is achieved. In this work a method is presented which uses a combination of statistical pattern recognition technique and damage detection algorithm in order to solve the problem of false alarms which refers to environmental and operational changes. The basic idea is to introduce reference states for the undamaged structure which are functions of the influences. The reference states are systematically found by using classification tools like the K-means or the Expectation-Maximization algorithm, respectively. The compensation of environmental influences is analyzed using a simulation model and a laboratory structure. A simplified beam model of an Offshore Wind Energy Plant (OWEP) with up to five input parameters is investigated. The model is simultaneously excited by loads with different probability distributions in the low frequency range. Simulation studies produce good results with only few false alarms due to statistical outliers. For verification purposes, experiments are performed on a laboratory structure that confirm the good simulation results.
Compensation of environmental influences for damage detection using classification techniques
2008
8 Seiten, 16 Quellen
Aufsatz (Konferenz)
Englisch
Combined damage detection techniques
Online Contents | 2003
|Damage Detection in Composite Materials Using Identification Techniques.
Online Contents | 1993
|Emerald Group Publishing | 1941