- Research area
Big data sensors and digitalisation for the offshore environment
University of Sheffield
- Research project
In-Service Inference of Distributed Blade Loading
- Lead supervisor
- PhD Student
- Supervisory Team
Professor Elizabeth Cross (Professor - Dept of Mechanical Engineering & EPSRC Innovation Fellow, University of Sheffield)
The load experienced by an offshore wind turbine blade in operation is significant and important to its operation. Ensuring structural integrity of the blades is a key concern when considering the operation and maintenance of the turbine. Direct measurement of the distributed load from the wind is, however, not possible with current sensing technology. Rather than developing new sensors, this project will seek to recover the load that the system has experienced from indirect measurement. Knowing that the dynamic motion of the blade is inherently linked to the experienced load, the dynamic system must be identified jointly with the unmeasured load. This project will seek to extend and develop the technology necessary for this prodigious challenge alongside validation of the proposed approach.
The problem of identification for systems with unknown and non-white-Gaussian loads is only growing in interest across the structural dynamics community, see [1-3]. The lead investigator of this project is at the forefront of these technologies, showing recent developments for joint input-state-parameter problems  and nonlinear input-state estimation .This project will build on those results to extend them to the case of distributed loads on the wind turbine blade.
Of particular interest will be the coversion of the linear coregionalisation model for multioutput Gaussian processes  into a state-space form for inclusion in a latent force modelling framework [7,8], where the dynamics is represented in modal coordinates. This methodology should allow identification of the dynamics of the wind turbine blade (its natural frequencies, damping ratios etc.) alongside the unmeasured distributed loads, which will be mapped back through the mode shapes to recover the spatial component of the load.
The validation of this approach will take place by assessing the performance of the proposed latent force modelling method in simultaneously recovering unsteady pressure distributions and aeroelastic responses of a model composite turbine blade in a wind tunnel at the University of Sheffield.
References & Further Reading
 Dertimanis, Vasilis K., et al. “Input-state-parameter estimation of structural systems from limited output information.” Mechanical Systems and Signal Processing 126 (2019): 711-746.
 Tatsis, K. E., et al. “A general substructure-based framework for input-state estimation using limited output measurements.” Mechanical Systems and Signal Processing 150 (2021): 107223.
 Petersen, Ø. W., O. Øiseth, and E. Lourens. “Wind load estimation and virtual sensing in long-span suspension bridges using physics-informed Gaussian process latent force models.” Mechanical Systems and Signal Processing 170 (2022): 108742.
 Rogers, T. J., K. Worden, and E. J. Cross. “On the application of Gaussian process latent force models for joint input-state-parameter estimation: With a view to Bayesian operational identification.” Mechanical Systems and Signal Processing 140 (2020): 106580.
 Rogers, Timothy J., Keith Worden, and Elizabeth J. Cross. “Bayesian joint input-state estimation for nonlinear systems.” Vibration 3.3 (2020): 281-303.
 Bonilla, Edwin V., Kian Chai, and Christopher Williams. “Multi-task Gaussian process prediction.” Advances in neural information processing systems 20 (2007).
 Alvarez, Mauricio, David Luengo, and Neil D. Lawrence. “Latent force models.” Artificial Intelligence and Statistics. PMLR, 2009.
 Hartikainen, Jouni, and Simo Sarkka. “Sequential inference for latent force models.” arXiv preprint arXiv:1202.3730 (2012).