Soft Ordering 1-D CNN to Estimate the Capacity Factor of Windfarms for Identifying the Age-Related Performance Degradation

Abstract

Wind energy plays a vital role in meeting the sustainable development goals set forth by the United Nations. Performance of wind energy farms degrades gradually with aging. For deriving maximum benefits from these capital-intensive projects, these degradation patten should be analyzed and understood. Variations in the capacity factor over the years could be an indication of the age-related degradation of the wind farms. In this study, we propose a novel data-driven model to estimate the capacity factor of wind farms, which could then be used to estimate its age-related performance decline. For this, a 1-dimensional convolutional neural network (1-D CNN) is developed with a soft ordering mechanism under this study. The model was optimized using Huber loss to counteract the effects of outliers in data. The developed model could perform very well in capturing the underlying dynamics in the data as evidenced by a normalized root mean squared error (NRMSE) of 0.102 and a mean absolute error (MAE) of 0.035 on the test dataset.