A novel and innovative solution addressing wind turbines’ main bearing failure predictions using SCADA data is presented. This methodology enables to cut setup times and has more flexible requirements when compared to the current predictive algorithms. The proposed solution is entirely unsupervised as it does not require the labeling of data through work orders logs. Results of interpretable algorithms, which are tailored to capture specific aspects of main bearing failures, are merged into a combined health status indicator making use of Ensemble Learning principles. Based on multiple specialized indicators, the interpretability of the results is greater compared to black-box solutions that try to address the problem with a single complex algorithm. The proposed methodology has been tested on a dataset covering more than two year of operations from two onshore wind farms, counting a total of 84 turbines. All four main bearing failures are anticipated at least one month of time in advance. Combining individual indicators into a composed one proved effective with regard to all the tracked metrics. Accuracy of 95.1%, precision of 24.5% and F1 score of 38.5% are obtained averaging the values across the two windfarms. The encouraging results, the unsupervised nature and the flexibility and scalability of the proposed solution are appealing, making it particularly attractive for any online monitoring system used on single wind farms as well as entire wind turbine fleets.

It is well known that each year the wind sector has profit losses due to wind turbine failures and operation and maintenance costs. Therefore, operations related to these actions are crucial for wind farm operators and linked companies. One of the key points for failure prediction on wind turbine using SCADA data is to select the optimal or near optimal set of inputs that can feed the failure prediction (prognosis) algorithm. Due to a high number of possible predictors (from tens to hundreds), the optimal set of inputs obtained by exhaustive-search algorithms is not viable in the majority of cases. In order to tackle this issue, show the viability of prognosis and select the best set of variables from more than 200 analogous variables recorded at intervals of 5 or 10 min by the wind farm’s SCADA, in this paper a thorough study of automatic input selection algorithms for wind turbine failure prediction is presented and an exhaustive-search-based quasi-optimal (QO) algorithm, which has been used as a reference, is proposed. In order to evaluate the performance, a k-NN classification algorithm is used. Results showed that the best automatic feature selection method in our case-study is the conditional mutual information (CMI), while the worst one is the mutual information feature selection (MIFS). Furthermore, the effect of the number of neighbours (k) is tested. Experiments demonstrate that k = 1 is the best option if the number of features is higher than 3. The experiments carried out in this work have been extracted from measures taken along an entire year and corresponding to gearbox and transmission systems of Fuhrländer wind turbines.

Product maintenance costs throughout the product’s lifetime can account for between 30–60% of total operating costs, making it necessary to implement maintenance strategies. This problem not only affects the economy but is also related to the impact on the environment, since breakdowns are also responsible for the delivery of greenhouse gases. Industrial maintenance is a set of measures of a technical-organizational nature whose purpose is to sustain the functionality of the equipment and guarantee an optimal state of the machines over time, with the aim of saving costs, extending the useful life of the machines, saving energy, maximising production and availability, ensuring the quality of the product obtained, providing job security for technicians, preserving the environment, and reducing emissions as much as possible. Machine learning techniques can be used to detect or predict faults in wind turbines. However, labelled data suffers from many problems in this application because alarms are usually not clearly associated with a specific fault, some labels are wrongly associated with a problem, and the imbalance between labels is evident. To avoid using labelled data, we investigate here the use of the clustering technique, more specifically K-means, and boxplot representations of the variables for a set of six different tests. Experimental results show that in some cases, the clustering and boxplot techniques allow us to determine outliers or identify erroneous behaviours of the wind turbines. These cases can then be investigated in detail by a specialist so that more efficient predictive maintenance can be carried out.

Double frequency tests are used for evaluating stator windings and analyzing the temperature. Likewise, signal injection on induction machines is used on sensorless motor control fields to find out the rotor position. Motor Current Signature Analysis (MCSA), which focuses on the spectral analysis of stator current, is the most widely used method for identifying faults in induction motors. Motor faults such as broken rotor bars, bearing damage and eccentricity of the rotor axis can be detected. However, the method presents some problems at low speed and low torque, mainly due to the proximity between the frequencies to be detected and the small amplitude of the resulting harmonics. This paper proposes the injection of an additional voltage into the machine being tested at a frequency different from the fundamental one, and then studying the resulting harmonics around the new frequencies appearing due to the composition between injected and main frequencies.

The goal of this paper is to develop, implement, and validate a methodology for wind turbines’ main bearing fault prediction based on an ensemble of an artificial neural network (normality model designed at turbine level) and an isolation forest (anomaly detection model designed at wind park level) algorithms trained only on SCADA data. The normal behavior and the anomalous samples of the wind turbines are identified and several interpretable indicators are proposed based on the predictions of these algorithms, to provide the wind park operators with understandable information with enough time to plan operations ahead and avoid unexpected costs. The stated methodology is validated in a real underproduction wind park composed by 18 wind turbines.

Background: Identifying the health status of wind turbines becomes critical to reduce the impact of failures on generation costs (between 25–35%). This is a time-consuming task since a human expert has to explore turbines individually. Methods: To optimize this process, we present a strategy based on Self Organizing Maps, clustering and a further grouping of turbines based on the centroids of their SOM clusters, generating groups of turbines that have similar behavior for subsystem failure. The human expert can diagnose the wind farm health by the analysis of a small each group sample. By introducing post-processing tools like Class panel graphs and Traffic lights panels, the conceptualization of the clusters is enhanced, providing additional information of what kind of real scenarios the clusters point out contributing to a better diagnosis. Results: The proposed approach has been tested in real wind farms with different characteristics (number of wind turbines, manufacturers, power, type of sensors, ...) and compared with classical clustering. Conclusions: Experimental results show that the states healthy, unhealthy and intermediate have been detected. Besides, the operational modes identified for each wind turbine overcome those obtained with classical clustering techniques capturing the intrinsic stationarity of the data.