Long-Term Multiple Sensor Monitoring of a Hybrid Tower Wind Turbine – Lessons Learned

Rupfle, Johannes; Wondra, Bernhard; Große, Christian U.

doi:10.1007/978-3-031-39109-5_55

Cited by 1 publication

(1 citation statement)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…General challenges of long-term multiple sensor monitoring of a hybrid tower wind turbine and lessons learned are detailed in [7], emphasizing seven years of experience in data processing and integration into digital twin environments for improved wind turbine sustainability and maintenance.…”

Section: Introductionmentioning

confidence: 99%

Operational Sensor Adjustment for Structural Health Monitoring Systems in Wind Turbines using Machine Learning

Rupfle,

Wondra,

Nowack

2024

Preprint

Self Cite

View full text Add to dashboard Cite

The digitalization in structural engineering has significantly amplified the importance of Structural Health Monitoring systems, especially in civil engineering and the renewable energy sector. A monitoring system allows for deep analysis of the structure's health state, can detect anomalies, and also allows for the determination of remaining useful life. Initially, after installation, a monitoring system needs adjustment so that it can be used for various analysis purposes. This paper presents three different methods to fully automatically adjust sensor offsets of tilt, acceleration, and strain data in wind turbine towers. In the first method, we make use of an initiated 360-degree nacelle rotation and its unbalanced rotor weight-induced tower bending. In the second method, uniaxial operational data is collected and binned to form a database for a univariate regression. A third method consists of a biaxial operational data set, allowing for a multivariate regression. The applications of the second and third approaches can be extended to any slender structure under environmental load. The presented workflow also allows for fully automated sensor orientation determination and fault detection due to long-term drift observation of sensor characteristics.

show abstract

Section: Introductionmentioning

confidence: 99%