Power Disturbance Monitoring through Techniques for Novelty Detection on Wind Power and Photovoltaic Generation.

Novelty detection is a statistical method that verifies new or unknown data, determines whether these data are inliers (within the norm) or outliers (outside the norm), and can be used, for example, in developing classification strategies in machine learning systems for industrial applications. To this end, two types of energy that have evolved over time are solar photovoltaic and wind power generation. Some organizations around the world have developed energy quality standards to avoid known electric disturbances; however, their detection is still a challenge. In this work, several techniques for novelty detection are implemented to detect different electric anomalies (disturbances), which are k-nearest neighbors, Gaussian mixture models, one-class support vector machines, self-organizing maps, stacked autoencoders, and isolation forests. These techniques are applied to signals from real power quality environments of renewable energy systems such as solar photovoltaic and wind power generation. The power disturbances that will be analyzed are considered in the standard IEEE-1159, such as sag, oscillatory transient, flicker, and a condition outside the standard attributed to meteorological conditions. The contribution of the work consists of the development of a methodology based on six techniques for novelty detection of power disturbances, under known and unknown conditions, over real signals in the power quality assessment. The merit of the methodology is a set of techniques that allow to obtain the best performance of each one under different conditions, which constitutes an important contribution to the renewable energy systems.

Diagnosis Methodology Based on Deep Feature Learning for Fault Identification in Metallic, Hybrid and Ceramic Bearings.

Scientific and technological advances in the field of rotatory electrical machinery are leading to an increased efficiency in those processes and systems in which they are involved. In addition, the consideration of advanced materials, such as hybrid or ceramic bearings, are of high interest towards high-performance rotary electromechanical actuators. Therefore, most of the diagnosis approaches for bearing fault detection are highly dependent of the bearing technology, commonly focused on the metallic bearings. Although the mechanical principles remain as the basis to analyze the characteristic patterns and effects related to the fault appearance, the quantitative response of the vibration pattern considering different bearing technology varies. In this regard, in this work a novel data-driven diagnosis methodology is proposed based on deep feature learning applied to the diagnosis and identification of bearing faults for different bearing technologies, such as metallic, hybrid and ceramic bearings, in electromechanical systems. The proposed methodology consists of three main stages: first, a deep learning-based model, supported by stacked autoencoder structures, is designed with the ability of self-adapting to the extraction of characteristic fault-related features from different signals that are processed in different domains. Second, in a feature fusion stage, information from different domains is integrated to increase the posterior discrimination capabilities during the condition assessment. Third, the bearing assessment is achieved by a simple softmax layer to compute the final classification results. The achieved results show that the proposed diagnosis methodology based on deep feature learning can be effectively applied to the diagnosis and identification of bearing faults for different bearing technologies, such as metallic, hybrid and ceramic bearings, in electromechanical systems. The proposed methodology is validated in front of two different electromechanical systems and the obtained results validate the adaptability and performance of the proposed approach to be considered as a part of the condition-monitoring strategies where different bearing technologies are involved.

Deep-Compact-Clustering Based Anomaly Detection Applied to Electromechanical Industrial Systems.

The rapid growth in the industrial sector has required the development of more productive and reliable machinery, and therefore, leads to complex systems. In this regard, the automatic detection of unknown events in machinery represents a greater challenge, since uncharacterized catastrophic faults can occur. However, the existing methods for anomaly detection present limitations when dealing with highly complex industrial systems. For that purpose, a novel fault diagnosis methodology is developed to face the anomaly detection. An unsupervised anomaly detection framework named deep-autoencoder-compact-clustering one-class support-vector machine (DAECC-OC-SVM) is presented, which aims to incorporate the advantages of automatically learnt representation by deep neural network to improved anomaly detection performance. The method combines the training of a deep-autoencoder with clustering compact model and a one-class support-vector-machine function-based outlier detection method. The addressed methodology is applied on a public rolling bearing faults experimental test bench and on multi-fault experimental test bench. The results show that the proposed methodology it is able to accurately to detect unknown defects, outperforming other state-of-the-art methods.

Deep-Learning-Based Methodology for Fault Diagnosis in Electromechanical Systems.

Fault diagnosis in manufacturing systems represents one of the most critical challenges dealing with condition-based monitoring in the recent era of smart manufacturing. In the current Industry 4.0 framework, maintenance strategies based on traditional data-driven fault diagnosis schemes require enhanced capabilities to be applied over modern production systems. In fact, the integration of multiple mechanical components, the consideration of multiple operating conditions, and the appearance of combined fault patterns due to eventual multi-fault scenarios lead to complex electromechanical systems requiring advanced monitoring strategies. In this regard, data fusion schemes supported with advanced deep learning technology represent a promising approach towards a big data paradigm using cloud-based software services. However, the deep learning models' structure and hyper-parameters selection represent the main limitation when applied. Thus, in this paper, a novel deep-learning-based methodology for fault diagnosis in electromechanical systems is presented. The main benefits of the proposed methodology are the easiness of application and high adaptability to available data. The methodology is supported by an unsupervised stacked auto-encoders and a supervised discriminant analysis.

Defect reconstruction by non-destructive testing with laser induced ultrasonic detection.

This work envisages a detailed study of two-dimensional defect localization and reconstruction, using laser generated ultrasound and its application as a remotely controlled non-destructive testing method. As an alternative to full ultrasonic or full optical approaches, we propose a hybrid configuration where ultrasound is generated by impact of laser pulses, while the detection is done with conventional transducers. We implement this approach for defect reconstruction in metallic elements and show that it combines advantages of both photonic and ultrasonic devices, reducing the drawbacks of both methods. We combine our experimental results with a high-resolution signal processing procedure based on the synthetic aperture focusing technique for the benefit of the final two-dimensional visualization of the defects.

Fully Noncontact Hybrid NDT for 3D Defect Reconstruction Using SAFT Algorithm and 2D Apodization Window.

Nondestructive testing of metallic objects that may contain embedded defects of different sizes is an important application in many industrial branches for quality control. Most of these techniques allow defect detection and its approximate localization, but few methods give enough information for its 3D reconstruction. Here we present a hybrid laser-transducer system that combines remote, laser-generated ultrasound excitation and noncontact ultrasonic transducer detection. This fully noncontact method allows access to scan areas on different object's faces and defect details from different angles/perspectives. This hybrid system can analyze the object's volume data and allows a 3D reconstruction image of the embedded defects. As a novelty for signal processing improvement, we use a 2D apodization window filtering technique, applied along with the synthetic aperture focusing algorithm, to remove the undesired effects due to side lobes and wide-angle reflections of propagating ultrasound waves, thus enhancing the resulting 3D image of the defect. Finally, we provide both qualitative and quantitative volumetric results that yield valuable information about defect location and size.

Laser Ultrasound Inspection Based on Wavelet Transform and Data Clustering for Defect Estimation in Metallic Samples.

Laser-generated ultrasound is a modern non-destructive testing technique. It has been investigated over recent years as an alternative to classical ultrasonic methods, mainly in industrial maintenance and quality control procedures. In this study, the detection and reconstruction of internal defects in a metallic sample is performed by means of a time-frequency analysis of ultrasonic waves generated by a laser-induced thermal mechanism. In the proposed methodology, we used wavelet transform due to its multi-resolution time frequency characteristics. In order to isolate and estimate the corresponding time of flight of eventual ultrasonic echoes related to internal defects, a density-based spatial clustering was applied to the resulting time frequency maps. Using the laser scan beam's position, the ultrasonic transducer's location and the echoes' arrival times were determined, the estimation of the defect's position was carried out afterwards. Finally, clustering algorithms were applied to the resulting geometric solutions from the set of the laser scan points which was proposed to obtain a two-dimensional projection of the defect outline over the scan plane. The study demonstrates that the proposed method of wavelet transform ultrasonic imaging can be effectively applied to detect and size internal defects without any reference information, which represents a valuable outcome for various applications in the industry.