Machine Learning Workshop 2024
RAVE Thematic Workshop on Machine Learning Applications for Offshore Wind Data Analysis
October 10th, 2024 10:00 - 16:00 (CEST)
Fraunhofer IWES (Am Fallturm 1, 28359 Bremen, Germany)
Hybrid event
Centered on the transformative power of machine learning in offshore wind data analysis, we aimed to bring together researchers, data scientists and machine learning enthusiast at any career stage, to explore together the latest techniques, their applications using RAVE data and exchange expertise.
Anish Venu (DNVGL) presents current applications of machine learning methods in a Keynote: RAVE Machine Learning end-to-end cycle: A complete overview on the RAVE ML Model covering following topics:
-
- Selection of data
- Cleaning, manipulation and preparation of training data
- Development of ML algorithm and comparison of results from different algorithms
- Validation & performance testing of the algorithm
- Issues faced/identification of issues
- Sensitivity analysis of the model
- Transfer of model from one turbine to another
- Flagging strategies
- Final deployment
- Other applications
- Future works
In case you have any questions regarding the RAVE machine learning workshop, please contact us via rave-forschungsarchiv@bsh.de.
Machine Learning Workshop 2024 - Presentations
10:00
"Welcome and Introduction"
Tanja Grießmann
Leibniz University Hannover
Presentation live and in oral form, therefore no slides.
10:15
"RAVE Machine Learning end-to-end cycle: A complete overview on the RAVE ML Model"
Anish Venu
DNV Energy Systems
11:15
"Wind Data Gap Filling and Localized Wind Profile Predictions via a Machine Learning Approach"
Farkhondeh Rouholahnejad and Martin Jonietz Alvarez
Fraunhofer IWES
13:00
“Neural Networks for Offshore Wind Turbine Converter Failure Prognosis"
Demitri Moros
EDF Energy UK & IDCORE
14:00
"Application of a NARX-Based Surrogate Model for Offshore Wind Turbine Structural Loads Prediction and Uncertainty Analysis"
Xu Ning
University of Bergen
15:00
"Assessment of a Deep Learning Surrogate Model for Wind Turbine Load Estimation Using RAVE Data"
Dexing Liu
University of Stuttgart