Ana María Peco Chacón, Isaac Segovia and Fausto Pedro García have been awarded with the Best Paper of the International Conference on Integrated Emerging Methods of Artificial Intelligence & Cloud Computing (IEMAICLOUD 2021) for the paper titled «False Alarm Detection in Wind Turbine Managmenet by K-Nearest Neighbors Models».
This work is focused on the fields of alarm analysis, control monitoring and machine learning. A real case study is presented with the dataset from a real wind turbine with the goal to detect false alarms. The novelty proposed in this paper is to use a K-nearest neighbor algorithm for the prediction and classification of false alarms. The holdout validation and k-fold validation have been evaluated for this algorithm. The highest accuracy is obtained by weighted K-nearest neighbor with 5-fold validation, with a value of 98,7%. The specificity and the sensitivity are 88,52% and 99,13%, respectively. These values imply that the methodology used is effective for false alarm detection and identification.