A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Intelligent estimation of the wake losses in wind farms : Artificial neural network estimation of the power of a wind farm considering the effect of wake losses
Master's thesis Renewable Energy ENE500 - University of Agder 2018 ; The transition from non-renewable to renewable energy production requires a detailed optimization and quantification of the generated power. The loss of power due to wake effect is a common problem for wind farms. The wake effect is the reduction of velocity and increase of turbulence in the wind flow downstream from a wind turbine. The wake effect is a complex multivariable phenomenon and its understanding iscapital forappropriate estimations of the power of a wind field and its turbines.This thesis builds an artificial neural network based on machine learning to model the performance of a single wind farm owned by WEICAN (Canada) taking into account the wake losses. Four different models have been considered. The first is not accounting for the wake losses; the second considers only the wake of the closest turbines; the third takes into account the wake in all the turbines; and the fourth provides all the data to the program in order to see what it can doon its own. The performance is evaluated using the mean absolute error, the root mean squared error and the normalized root mean square error.The best results areobtained using the third model, hence showing that the wake loss is significant and must be considered in the model. It is proved that with the appropriate input variables, an artificial neural network can predict the power of a wind farm accounting for the wake losses. The best performance of the artificial neural network is obtained for wind speeds up to 14 m/s.
Intelligent estimation of the wake losses in wind farms : Artificial neural network estimation of the power of a wind farm considering the effect of wake losses
Master's thesis Renewable Energy ENE500 - University of Agder 2018 ; The transition from non-renewable to renewable energy production requires a detailed optimization and quantification of the generated power. The loss of power due to wake effect is a common problem for wind farms. The wake effect is the reduction of velocity and increase of turbulence in the wind flow downstream from a wind turbine. The wake effect is a complex multivariable phenomenon and its understanding iscapital forappropriate estimations of the power of a wind field and its turbines.This thesis builds an artificial neural network based on machine learning to model the performance of a single wind farm owned by WEICAN (Canada) taking into account the wake losses. Four different models have been considered. The first is not accounting for the wake losses; the second considers only the wake of the closest turbines; the third takes into account the wake in all the turbines; and the fourth provides all the data to the program in order to see what it can doon its own. The performance is evaluated using the mean absolute error, the root mean squared error and the normalized root mean square error.The best results areobtained using the third model, hence showing that the wake loss is significant and must be considered in the model. It is proved that with the appropriate input variables, an artificial neural network can predict the power of a wind farm accounting for the wake losses. The best performance of the artificial neural network is obtained for wind speeds up to 14 m/s.
Intelligent estimation of the wake losses in wind farms : Artificial neural network estimation of the power of a wind farm considering the effect of wake losses
Tarragó Aymerich, Martí (author)
2018-01-01
61 p.
Theses
Electronic Resource
English
DDC:
690
Offshore Wind Farm Layout Optimisation Considering Wake Effect and Power Losses
| DOAJ | 2023
Adaptive Frequency Responsive Control for Wind Farm Considering Wake Interaction
| DOAJ | 2021
Graph network heterogeneity predicts interplant wake losses
| American Institute of Physics | 2024