EXPERIMENTAL VALIDATION OF A STRUCTURAL HEALTH MONITORING METHODOLOGY: PART II. NOVELTY DETECTION ON A GNAT AIRCRAFT

2003 ◽  
Vol 259 (2) ◽  
pp. 345-363 ◽  
Author(s):  
G. MANSON ◽  
K. WORDEN ◽  
D. ALLMAN
Keyword(s):  
Structural Health Monitoring ◽  
Health Monitoring ◽  
Experimental Validation ◽  
Novelty Detection ◽  
Structural Health

scholarly journals Structural Health Monitoring under Nonlinear Environmental or Operational Influences

2014 ◽  
Vol 2014 ◽  
pp. 1-9 ◽  
Author(s):  
Jyrki Kullaa
Keyword(s):  
Structural Health Monitoring ◽  
Health Monitoring ◽  
Linear Models ◽  
Nonlinear Models ◽  
Novelty Detection ◽  
Minimum Mean Square Error ◽  
Simulated Data ◽  
Principal Component ◽  
Gaussian Mixture ◽  
Structural Health

Vibration-based structural health monitoring is based on detecting changes in the dynamic characteristics of the structure. It is well known that environmental or operational variations can also have an influence on the vibration properties. If these effects are not taken into account, they can result in false indications of damage. If the environmental or operational variations cause nonlinear effects, they can be compensated using a Gaussian mixture model (GMM) without the measurement of the underlying variables. The number of Gaussian components can also be estimated. For the local linear components, minimum mean square error (MMSE) estimation is applied to eliminate the environmental or operational influences. Damage is detected from the residuals after applying principal component analysis (PCA). Control charts are used for novelty detection. The proposed approach is validated using simulated data and the identified lowest natural frequencies of the Z24 Bridge under temperature variation. Nonlinear models are most effective if the data dimensionality is low. On the other hand, linear models often outperform nonlinear models for high-dimensional data.

Vibration-Based Structural Health Monitoring of a Simulated Beam with a Breathing Crack

2013 ◽  
Vol 569-570 ◽  
pp. 1093-1100 ◽  
Author(s):  
Jyrki Kullaa ◽  
Kari Santaoja ◽  
Anthony Eymery
Keyword(s):  
Structural Health Monitoring ◽  
Damage Detection ◽  
Crack Length ◽  
Health Monitoring ◽  
Early Stage ◽  
Novelty Detection ◽  
Principal Component ◽  
Extreme Value Statistics ◽  
Ratio Test ◽  
Structural Health

Cracking is a common type of failure in machines and structures. Cracks must be detected at an early stage before catastrophic failure. In structural health monitoring, changes in the vibration characteristics of the structure can be utilized in damage detection. A fatigue crack with alternating contact and non-contact phases results in a non-linear behaviour. This type of damage was simulated with a finite element model of a simply supported beam. The structure was monitored with a sensor array measuring transverse accelerations under random excitation. The objective was to determine the smallest crack length that can be detected. The effect of the sensor locations was also studied. Damage detection was performed using the generalized likelihood ratio test (GLRT) in time domain followed by principal component analysis (PCA). Extreme value statistics (EVS) were used for novelty detection. It was found that a crack in the bottom of the midspan could be detected once the crack length exceeded 10% of the beam height. The crack was correctly localized using the monitoring data.

Structural Health Monitoring of Composite Material Typical of Wind Turbine Blades by Novelty Detection on Vibration Response

2012 ◽  
Vol 518 ◽  
pp. 319-327
Author(s):  
Nikolaos Dervilis ◽  
R. Barthorpe ◽  
Wieslaw Jerzy Staszewski ◽  
Keith Worden
Keyword(s):  
Composite Material ◽  
Structural Health Monitoring ◽  
Damage Detection ◽  
Wind Turbine ◽  
Wind Turbines ◽  
Health Monitoring ◽  
Novelty Detection ◽  
Turbine Blades ◽  
Wind Turbine Blades ◽  
Structural Health

New generations of offshore wind turbines are playing a leading role in the energy arena. One of the target challenges is to achieve reliable Structural Health Monitoring (SHM) of the blades. Fault detection at the early stage is a vital issue for the structural and economical success of the large wind turbines. In this study, experimental measurements of Frequency Response Functions (FRFs) are used and identification of mode shapes and natural frequencies is accomplished via an LMS system. Novelty detection is introduced as a robust statistical method for low-level damage detection which has not yet been widely used in SHM of composite blades. Fault diagnosis of wind turbine blades is a challenge due to their composite material, dimensions, aerodynamic nature and environmental conditions. The novelty approach combined with vibration measurements introduces an online condition monitoring method. This paper presents the outcomes of a scheme for damage detection of carbon fibre material in which novelty detection approaches are applied to FRF measurements. The approach is demonstrated for a stiffened composite plate subject to incremental levels of impact damage.

Optimized Actuator/Sensor Combinations for Structural Health Monitoring: Simulation and Experimental Validation

2015 ◽  
Author(s):  
SRIDARAN VENKAT ◽  
C. BOLLER ◽  
N.B. RAVI ◽  
N. CHAKRABORTY ◽  
G.S. KAMALAKAR ◽  
...  
Keyword(s):  
Structural Health Monitoring ◽  
Health Monitoring ◽  
Experimental Validation ◽  
Structural Health

Some Recent Developments in Structural Health Monitoring

2012 ◽  
Vol 518 ◽  
pp. 298-318 ◽  
Author(s):  
R.J. Barthorpe ◽  
E.J. Cross ◽  
E. Papatheou ◽  
Keith Worden
Keyword(s):  
Structural Health Monitoring ◽  
Unsupervised Learning ◽  
Supervised Learning ◽  
Health Monitoring ◽  
Research Group ◽  
Novelty Detection ◽  
Structural Health ◽  
Recent Developments ◽  
Damage States ◽  
The University

This paper is concerned with reporting some recent developments in Structural Health Monitoring (SHM) research conducted within the Dynamics Research Group at the University of Sheffield. The particular developments discussed are concerned with arguably the two main problems facing data-based approaches to SHM, namely: how to obtain data from damage states of a structure for supervised learning and how to remove environmental and operational effects from data when unsupervised learning (novelty detection) is indicated.

Sign in / Sign up

Export Citation Format

Share Document