Performance analysis and anomaly detection in wind turbines based on neural networks and principal component analysis

Abstract

This paper proposes an approach for maintenance management of wind turbines based on their life. The proposed approach uses performance analysis and anomaly detection (PAAD) which can detect anomalies and point out the origin of the detected anomalies. This PAAD algorithm utilizes neural network (NN) technique in order to detect anomalies in the performance of the wind turbine (system layer), and then applies principal component analysis (PCA) technique to uncover the root of the detected anomalies (component layer). To validate the accuracy of the proposed algorithm, SCADA data obtained from online condition monitoring of a wind turbine are utilized. The results demonstrate that the proposed PAAD algorithm has the capability of exposing the cause of the anomalies. Reducing time and cost of maintenance and increasing availability and in return profits in form of savings are some of the benefits of the proposed PAAD algorithm.This paper proposes an approach for maintenance management of wind turbines based on their life. The proposed approach uses performance analysis and anomaly detection (PAAD) which can detect anomalies and point out the origin of the detected anomalies. This PAAD algorithm utilizes neural network (NN) technique in order to detect anomalies in the performance of the wind turbine (system layer), and then applies principal component analysis (PCA) technique to uncover the root of the detected anomalies (component layer). To validate the accuracy of the proposed algorithm, SCADA data obtained from online condition monitoring of a wind turbine are utilized. The results demonstrate that the proposed PAAD algorithm has the capability of exposing the cause of the anomalies. Reducing time and cost of maintenance and increasing availability and in return profits in form of savings are some of the benefits of the proposed PAAD algorithm.