SAR Surface Wind Estimation and Extrapolation at Turbine Hub Height with Machine Learning for Offshore Wind Farm Siting

2018 ◽

Author(s):

Alexios Koltsidopoulos Papatzimos ◽

Tariq Dawood ◽

Philipp R. Thies

Keyword(s):

Machine Learning ◽

Wind Farm ◽

Service Providers ◽

Learning Algorithms ◽

Machine Learning Algorithms ◽

Offshore Wind ◽

Sensor Data ◽

Support Vector ◽

Offshore Wind Turbine ◽

Offshore Wind Farm

Offshore wind assets have reached multi-GW scale and additional capacity is being installed and developed. To achieve demanding cost of energy targets, awarded by competitive auctions, the operation and maintenance (O&M) of these assets has to become increasingly efficient, whilst ensuring compliance and effectiveness. Existing offshore wind farm assets generate a significant amount of inhomogeneous data related to O&M processes. These data contain rich information about the condition of the assets, which is rarely fully utilized by the operators and service providers. Academic and industrial research and development efforts have led to a suite of tools trying to apply sensor data and build machine learning models to diagnose, trend and predict component failures. This study presents a decision support framework incorporating a range of different supervised and un-supervised learning algorithms. The aim is to provide guidance for asset owners on how to select the most relevant datasets, apply and choose the different machine learning algorithms and how to integrate the data stream with daily maintenance procedures. The presented methodology is tested on a real case example of an offshore wind turbine gearbox replacement at Teesside offshore wind farm. The study uses k-nearest neighbour (kNN) and support vector machine (SVM) algorithms to detect the fault using supervisory control and data acquisition (SCADA) data and an autoregressive model for the vibration data of the condition monitoring system (CMS). The implementation of all the algorithms has resulted in an accuracy higher than 94%. The results of this paper will be of interest to offshore wind farm developers and operators to streamline and optimize their O&M planning activities for their assets and reduce the associated costs.

Download Full-text

Sea surface wind measurement over offshore wind farm using TerraSAR-X data

10.1117/12.910331 ◽

2011 ◽

Author(s):

Xiao-Ming Li ◽

Susanne Lehner ◽

Stephan Brusch ◽

Yong-Zheng Ren

Keyword(s):

Wind Farm ◽

Surface Wind ◽

Sea Surface ◽

Offshore Wind ◽

Offshore Wind Farm ◽

Wind Measurement ◽

Sea Surface Wind

Download Full-text

Big data driven multi-objective predictions for offshore wind farm based on machine learning algorithms

Energy ◽

10.1016/j.energy.2019.07.034 ◽

2019 ◽

Vol 186 ◽

pp. 115704 ◽

Cited By ~ 7

Author(s):

Xiuxing Yin ◽

Xiaowei Zhao

Keyword(s):

Machine Learning ◽

Big Data ◽

Wind Farm ◽

Learning Algorithms ◽

Machine Learning Algorithms ◽

Offshore Wind ◽

Data Driven ◽

Offshore Wind Farm ◽

Multi Objective

Download Full-text

Satellite Observations for Better Characterization of Sea Surface Wind Field and Offshore Wind Energy Resource Assessment

10.4043/31316-ms ◽

2021 ◽

Author(s):

Mauricio Fragoso ◽

Louis De Montera ◽

Romain Husson ◽

Henrick Berger ◽

Pascal Appelghem ◽

...

Keyword(s):

Machine Learning ◽

Numerical Model ◽

Wind Power ◽

Wind Farm ◽

Learning Algorithm ◽

Surface Wind ◽

Energy Resource ◽

Resource Assessment ◽

Offshore Wind ◽

Sar Data

Abstract This paper presents a method to generate maps of offshore wind power at turbine hub height from spaceborne Synthetic Aperture Radar (SAR) data. Two techniques based on machine learning are presented. The first can be trained with metocean buoys and the second one, more precise, requires on-site profiling Lidars. If Lidars are not available, SAR surface winds at 10m are improved with machine learning. They are then extrapolated at 40m with a classical power law, and then at higher altitudes with an atmospheric numerical model. If profiling Lidars are available, parameters from the numerical model are added as input to the machine learning algorithm and the training is performed directly at turbine hub height with the Lidar data. Once the wind at turbine hub height is obtained, the wind power is then calculated using a Weibull distribution. The resulting maps are compared with the outputs of the numerical model. The maps based on SAR data provide a much higher level of detail and a better estimation of the coastal gradient, which is important to optimize wind farm siting and estimate the potential energy production. The accuracy of the wind power is found to be in the range ±5% compared to the Lidars.

Download Full-text