Abstract
The new machine learning techniques offer improvements in the monitoring control of wind farms. Monitoring systems regularly report wind turbine conditions, this information can be used by classification algorithms to detect possible failures. The objective of this research is the classification of different types of alarms by means of different K-nearest neighbor algorithms. The approach is applied to an actual dataset of a wind turbine. The comparison between different validation techniques indicates that the best accuracy is 98.7% with weighted K-nearest neighbor algorithm and 10-fold crossvalidation. The results show that it can predict and classify alarms of the specification type of environmental states and conditions, with true positive rates of 89.4% and 75.9%, respectively. These results demonstrate that the proposed algorithms are efficient to classify this type of alarm. For future research, the methodology can be used with other algorithms and analyze their correlation with specific supervisory control and data acquisition signals.
Sign-in/Register to access full text options 
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 callMore From: Non-Destructive Testing and Condition Monitoring Techniques in Wind Energy
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
