Evaluation of three potential machine learning algorithms for predicting the velocity and turbulence intensity of a wind turbine wake

This article discusses the application of widely used, state of the art, wake models, focusing on the Ainslie [1], Katic [2] and Larsen [3] models, breaking these down and explaining the individual, integral components. Models used to predict the turbulence intensity within the wake are also explained. Measured data are subsequently used to validate these wake and turbulence models, showing acceptable results for the prediction of velocity deficit within the wake, wake width and wake shape. Results also highlight the validity of the analysed turbulence models. The paper describes the problems encountered when using measured data to validate wake models and concludes by outlining subsequent work which could be carried out to further validate these models.

Download Full-text

Wind Turbine Blade Damage Detection Using Supervised Machine Learning Algorithms

Journal of Vibration and Acoustics ◽

10.1115/1.4036951 ◽

2017 ◽

Vol 139 (6) ◽

Cited By ~ 14

Author(s):

Taylor Regan ◽

Christopher Beale ◽

Murat Inalpolat

Keyword(s):

Machine Learning ◽

Damage Detection ◽

Wind Turbine ◽

Health Monitoring ◽

Learning Algorithms ◽

Turbine Blades ◽

Machine Learning Algorithms ◽

Test Rig ◽

Wind Turbine Blades ◽

Baseline Characteristics

Wind turbine blades undergo high operational loads, experience variable environmental conditions, and are susceptible to failure due to defects, fatigue, and weather-induced damage. These large-scale composite structures are fundamentally enclosed acoustic cavities and currently have limited, if any, structural health monitoring (SHM) in place. A novel acoustics-based structural sensing and health monitoring technique is developed, requiring efficient algorithms for operational damage detection of cavity structures. This paper describes the selection of a set of statistical features for acoustics-based damage detection of enclosed cavities, such as wind turbine blades, as well as a systematic approach used in the identification of competent machine learning algorithms. Logistic regression (LR) and support vector machine (SVM) methods are identified and used with optimal feature selection for decision-making via binary classification algorithms. A laboratory-scale wind turbine with hollow composite blades was built for damage detection studies. This test rig allows for testing of stationary or rotating blades, of which time and frequency domain information can be collected to establish baseline characteristics. The test rig can then be used to observe any deviations from the baseline characteristics. An external microphone attached to the tower will be utilized to monitor blade health while blades are internally ensonified by wireless speakers. An initial test campaign with healthy and damaged blade specimens is carried out to arrive at several conclusions on the detectability and feature extraction capabilities required for damage detection.

Download Full-text

Volumetric Lidar Scanning of Wind Turbine Wakes under Convective and Neutral Atmospheric Stability Regimes

Journal of Atmospheric and Oceanic Technology ◽

10.1175/jtech-d-13-00252.1 ◽

2014 ◽

Vol 31 (10) ◽

pp. 2035-2048 ◽

Cited By ~ 63

Author(s):

Giacomo Valerio Iungo ◽

Fernando Porté-Agel

Keyword(s):

Thermal Stability ◽

Wind Turbine ◽

Wind Power ◽

Wind Velocity ◽

Turbulence Intensity ◽

Atmospheric Stability ◽

Power Harvesting ◽

Turbine Wake ◽

Wind Turbine Wake

Abstract Optimization of a wind farm’s layout is a strategic task to reduce wake effects on downstream turbines, thus maximizing wind power harvesting. However, downstream evolution and recovery of each wind turbine wake are strongly affected by the characteristics of the incoming atmospheric boundary layer (ABL) flow, such as the vertical profiles of the mean wind velocity and the turbulence intensity, which are in turn affected by the ABL thermal stability. Therefore, the characterization of the variability of wind turbine wakes under different ABL stability regimes becomes fundamental to better predict wind power harvesting and to improve wind farm efficiency. To this aim, wind velocity measurements of the wake produced by a 2-MW Enercon E-70 wind turbine were performed with three scanning Doppler wind lidars. One lidar was devoted to the characterization of the incoming wind—in particular, wind velocity, shear, and turbulence intensity at the height of the rotor disc. The other two lidars performed volumetric scans of the wind turbine wake under different atmospheric conditions. Through the evaluation of the minimum wake velocity deficit as a function of the downstream distance, it is shown that the ABL stability regime has a significant effect on the wake evolution; in particular, the wake recovers faster under convective conditions. This result suggests that atmospheric inflow conditions, and particularly thermal stability, should be considered for improved wake models and predictions of wind power harvesting.

Download Full-text

Data-driven wind turbine wake modeling via probabilistic machine learning

Neural Computing and Applications ◽

10.1007/s00521-021-06799-6 ◽

2022 ◽

Author(s):

S. Ashwin Renganathan ◽

Romit Maulik ◽

Stefano Letizia ◽

Giacomo Valerio Iungo

Keyword(s):

Machine Learning ◽

Wind Turbine ◽

Data Driven ◽

Probabilistic Machine Learning ◽

Turbine Wake ◽

Wake Modeling ◽

Wind Turbine Wake

Download Full-text

AD/RANS Simulations of Wind Turbine Wake Flow Employing the RSM Turbulence Model: Impact of Isotropic and Anisotropic Inflow Conditions

Energies ◽

10.3390/en12214026 ◽

2019 ◽

Vol 12 (21) ◽

pp. 4026 ◽

Cited By ~ 1

Author(s):

Tian ◽

Song ◽

Zhao ◽

Shen ◽

Wang

Keyword(s):

Wind Turbine ◽

Turbulence Intensity ◽

Isotropic Turbulence ◽

Sensitivity Study ◽

Wake Flow ◽

Wake Region ◽

Anisotropic Turbulence ◽

Wide Range ◽

Turbine Wake ◽

Wind Turbine Wake

The Reynolds-averaged Navier–Stokes (RANS)-based generalized actuator disc method along with the Reynolds stress model (AD/RANS_RSM) is assessed for wind turbine wake simulation. The evaluation is based on validations with four sets of experiments for four horizontal-axis wind turbines with different geometrical characteristics operating in a wide range of wind conditions. Additionally, sensitivity studies on inflow profiles (representing isotropic and anisotropic turbulence) for predicting wake effects are carried out. The focus is on the prediction of the evolution of wake flow in terms of wind velocity and turbulence intensity. Comparisons between the computational results and the measurements demonstrate that in the near and transition wake region with strong anisotropic turbulence, the AD/RANS_RSM methodology exhibits a reasonably good match with all the experimental data sets; however, in the far wake region that is characterized by isotropic turbulence, the AD/RANS_RSM predicts the wake velocity quite accurately but appears to over-estimate the wake turbulence level. While the introduction of the overall turbulence intensity is found to give an improved agreement with the experiments. The performed sensitivity study proves that the anisotropic inflow condition is recommended as the profile of choice to represent the incoming wind flow.

Download Full-text