Gaussian process models for mitigation of operational variability in the structural health monitoring of wind turbines

Abstract The analysis presented in this work relates to the quantification of the effect of a selected set of measured Environmental and Operational Parameters (EOPs) on the dynamic properties of low and high frequency vibration, in the context of a vibration monitoring campaign implemented on the blade of an operating wind turbine. To this end, a Gaussian Process (GP) time-series modelling approach is adopted, in which the coefficients of a time-series model are driven by a Gaussian Process Regression on the selected EOPs. The properties of the data acquisition system allow to evaluate low and high frequency dynamics, the former associated with the structural dynamics of the blade, and the latter with the wave transmission properties of the material, assessed with the help of an electro-mechanical actuator installed on the blade. In this form, a multi-temporal-scale approach is adopted here, where a GP Linear Parameter Varying Auto-Regressive model is selected to represent low frequency (structural) dynamics, while in parallel a GP Continuous Wavelet Transform model is used to represent high frequency dynamics (associated with wave transmission properties in the material). In both cases the blade is considered in its healthy state as well as in various operational regimes, including idle, and rotating at two different set points. As a result, it is demonstrated that GP time-series modelling succeeds in evaluating and isolating the influence of different EOPs in the features of the vibration response of the wind turbine blade, and at the same time, normalize their effects to enhance the detectability of damage.

[1]  Keith Worden,et al.  On switching response surface models, with applications to the structural health monitoring of bridges , 2018 .

[2]  S. Mallat.  A wavelet tour of signal processing , 1998 .

[3]  Keith Worden,et al.  Cointegration and why it works for SHM , 2012 .

[4]  David A. Nix,et al.  Vibration–based structural damage identification , 2001, Philosophical Transactions of the Royal Society of London. Series A: Mathematical, Physical and Engineering Sciences.

[5]  Spilios D. Fassois,et al.  Parametric time-domain methods for non-stationary random vibration modelling and analysis — A critical survey and comparison , 2006 .

[6]  Eleni Chatzi,et al.  A Data-Driven Diagnostic Framework for Wind Turbine Structures: A Holistic Approach , 2017, Sensors.

[7]  Edwin Reynders,et al.  Output-only structural health monitoring in changing environmental conditions by means of nonlinear system identification , 2014 .

[8]  Hoon Sohn,et al.  Effects of environmental and operational variability on structural health monitoring , 2007, Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences.

[9]  Tomasz Barszcz,et al.  Nonlinear Cointegration Approach for Condition Monitoring of Wind Turbines , 2015 .

[10]  Rune Brincker,et al.  A new scenario-based approach to damage detection using operational modal parameter estimates , 2017 .

[11]  Keith Worden,et al.  A multiresolution approach to cointegration for enhanced SHM of structures under varying conditions – An exploratory study , 2014 .

[12]  G. De Roeck,et al.  Vibration based Structural Health Monitoring using output-only measurements under changing environment , 2008 .

[13]  Jyrki Kullaa,et al.  Distinguishing between sensor fault, structural damage, and environmental or operational effects in structural health monitoring , 2011 .

[14]  Spilios D. Fassois,et al.  In-Operation Wind Turbine Modal Analysis via LPV-VAR Modeling , 2017 .

[15]  Babak Moaveni,et al.  Environmental effects on the identified natural frequencies of the Dowling Hall Footbridge , 2011 .

[16]  Juan Ignacio Godino-Llorente,et al.  Feature Extraction From Parametric Time–Frequency Representations for Heart Murmur Detection , 2010, Annals of Biomedical Engineering.

[17]  Carlo Novara,et al.  Linear Parameter-Varying System Identification , 2012 .

[18]  Spilios D. Fassois,et al.  Damage/fault diagnosis in an operating wind turbine under uncertainty via a vibration response Gaussian mixture random coefficient model based framework , 2017 .

[19]  Spilios D. Fassois,et al.  Functionally Pooled models for the global identification of stochastic systems under different pseudo-static operating conditions , 2016 .

[20]  Spilios D. Fassois,et al.  Parametric identification of a time-varying structure based on vector vibration response measurements ☆ , 2009 .

[21]  James M. W. Brownjohn,et al.  Gaussian Process Time-Series Models for Structures under Operational Variability , 2017, Front. Built Environ..

[22]  Dmitri Tcherniak,et al.  Active vibration-based structural health monitoring system for wind turbine blade: Demonstration on an operating Vestas V27 wind turbine , 2017 .

[23]  Eleni Chatzi,et al.  Polynomial chaos expansion models for the monitoring of structures under operational variability , 2016 .

[24]  J. Rafiee,et al.  Application of mother wavelet functions for automatic gear and bearing fault diagnosis , 2010, Expert Syst. Appl..

[25]  Gaëtan Kerschen,et al.  Structural damage diagnosis under varying environmental conditions—Part I: A linear analysis , 2005 .

[26]  Spilios D. Fassois,et al.  Gaussian Mixture Random Coefficient model based framework for SHM in structures with time–dependent dynamics under uncertainty , 2017 .

[27]  S. Firstov,et al.  Universal temperature dependence of Young’s modulus , 2019, Metal Powder Report.

[28]  Dmitri Tcherniak,et al.  Active vibration-based SHM system: demonstration on an operating Vestas V27 wind turbine , 2016 .

[29]  Carlo Novara,et al.  Linear Parameter-Varying System Identification: New Developments and Trends , 2011 .

[30]  Marek Krawczuk,et al.  Improvement of damage detection methods based on experimental modal parameters , 2011 .

[31]  Keith Worden,et al.  A nonlinear cointegration approach with applications to structural health monitoring , 2016 .

[32]  Spilios D. Fassois,et al.  Stationary and non-stationary random vibration modelling and analysis for an operating wind turbine , 2014 .

[33]  Eleni Chatzi,et al.  Sensitivity driven robust vibration-based damage diagnosis under uncertainty through hierarchical Bayes time-series representations , 2017 .

[34]  Charles R. Farrar,et al.  Novelty detection in a changing environment: Regression and interpolation approaches , 2002 .

[35]  Eleni Chatzi,et al.  Metamodeling of dynamic nonlinear structural systems through polynomial chaos NARX models , 2015 .