Articles | Volume 6, issue 5
Wind Energ. Sci., 6, 1107–1116, 2021
https://doi.org/10.5194/wes-6-1107-2021
Wind Energ. Sci., 6, 1107–1116, 2021
https://doi.org/10.5194/wes-6-1107-2021

Research article 08 Sep 2021

Research article | 08 Sep 2021

On sensor optimisation for structural health monitoring robust to environmental variations

Tingna Wang et al.

Related subject area

Operation, condition monitoring, and maintenance
Effect of individual blade pitch angle misalignment on the remaining useful life of wind turbines
Matthias Saathoff, Malo Rosemeier, Thorsten Kleinselbeck, and Bente Rathmann
Wind Energ. Sci., 6, 1079–1087, https://doi.org/10.5194/wes-6-1079-2021,https://doi.org/10.5194/wes-6-1079-2021, 2021
Short summary

Cited articles

Barthorpe, R. J. and Worden, K.: Emerging trends in optimal structural health monitoring system design: From sensor placement to system evaluation, Journal of Sensor and Actuator Networks, 9, 31–51, 2020. a
Bull, L. A., Worden, K., Fuentes, R., Manson, G., Cross, E. J., and Dervilis, N.: Outlier ensembles: A robust method for damage detection and unsupervised feature extraction from high-dimensional data, J. Sound Vib., 453, 126–150, 2019. a
Cross, E. J., Manson, G., Worden, K., and Pierce, S. G.: Features for damage detection with insensitivity to environmental and operational variations, P. Roy. Soc. A-Math. Phy., 468, 4098–4122, 2012. a, b
Deraemaeker, A. and Worden, K.: A comparison of linear approaches to filter out environmental effects in structural health monitoring, Mech. Syst. Signal Pr., 105, 1–15, 2018. a, b, c, d
Eshghi, A. T., Lee, S., Jung, H., and Wang, P.: Design of structural monitoring sensor network using surrogate modeling of stochastic sensor signal, Mech. Syst. Signal Pr., 133, 106280, https://doi.org/10.1016/j.ymssp.2019.106280, 2019. a
Download
Short summary
This paper illustrates two sensor placement optimisation techniques designed for damage detection while taking into account temperature effects. A case study of a glider wing shows that, compared to the normalised method using the temperature label, the linear method that did not require temperature labels provided features that were less sensitive to damage. However, it is cheaper and more convenient to extract temperature-robust features in practical engineering.