Abstract
A Kernel Entropy Method and its Application in Monitoring and Assessment of Wind Turbine Degradation Performance
Highlights
With the intensification of the energy crisis, wind energy has become an important source of renewable energy and has gradually played a significant role in the global energy mix
Based on the error results presented in Tab. 1, the mean absolute error (MAE) and mean relative error (MRE) of the kernel entropy partial least squares (KEPLS) model are relatively small, suggesting that the output of the model can be directly compared with the actual temperature to evaluate whether the wind turbine (WT) is in an abnormal operating state
This paper proposed an algorithm for monitoring and assessment of wind turbine degradation performance based on the kernel entropy method
Summary
With the intensification of the energy crisis, wind energy has become an important source of renewable energy and has gradually played a significant role in the global energy mix. A set of lowdimensional variables containing important features that summarize the information carried by high-dimensional data can be constructed using multivariate statistical monitoring methods, such as principal component analysis (PCA), partial least squares (PLS) approach, independent component analysis (ICA), and other related algorithms [15]. To this end, considering the non-linear and non-stationary characteristics of SCADA data, a new fault prediction method based on kernel entropy PLS (KEPLS) is proposed. To better monitor the condition of WT components and predict future faults, this paper proposes a KEPLS predictive model that extracts multi-scale information more effectively and analyzes the residual error to achieve more accurate fault prediction.
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call