Miguel Delgado-Prieto

By Miguel Delgado Prieto, Jesús A. Cariño Corrales, Daniel Zurita Millán, Marta Millán Gonzalvez, Juan A. Ortega Redondo and René de J. Romero Troncoso

Dealing with industrial applications, the implementation of condition monitoring schemes must overcome a critical limitation, that is, the lack of a priori information about fault patterns of the system under analysis. Indeed, classical diagnosis schemes, in general, outdo the membership probability of a measure in regard to predefined operating scenarios. However, dealing with noncharacterized systems, the knowledge about faulty operating scenarios is limited and, consequently, the diagnosis performance is insufficient. In this context, the novelty detection framework plays an essential role for monitoring systems in which the information about different operating scenarios is initially unavailable or restricted. The novelty detection approach begins with the assumption that only data corresponding to the healthy operation of the system under analysis is available. Thus, the challenge is to detect and learn additional scenarios during the operation of the system in order to complement the information obtained by the diagnosis scheme. This work has two main objectives: first, the presentation of novelty detection as the current trend toward the new paradigm of industrial condition monitoring and, second, the introduction to its applicability by means of analyses of different novelty detection strategies over a real industrial system based on rotatory machinery.

Part of the book: Fault Diagnosis and Detection

By Hossam Selim, Fernando Piñal Moctezuma, Miguel Delgado Prieto, José Francisco Trull, Luis Romeral Martínez and Crina Cojocaru

Laser-generated ultrasound represents an interesting nondestructive testing technique that is being investigated in the last years as performative alternative to classical ultrasonic-based approaches. The greatest difficulty in analyzing the acoustic emission response is that an in-depth knowledge of how acoustic waves propagate through the tested composite is required. In this regard, different signal processing approaches are being applied in order to assess the significance of features extracted from the resulting analysis. In this study, the detection capabilities of internal defects in a metallic sample are proposed to be studied by means of the time-frequency analysis of the ultrasonic waves resulting from laser-induced thermal mechanism. In the proposed study, the use of the wavelet transform considering different wavelet variants is considered due to its multi-resolution time-frequency characteristics. Also, a significant time-frequency technique widely applied in other fields of research is applied, the synchrosqueezed transform.

Part of the book: Wavelet Transform and Complexity

By Joan Valls Pérez, Mayra Ramírez Chávez, Miguel Delgado-Prieto and Luis Romeral Martínez

Over the past few years, the anomaly detection problem has been intensively researched within different areas and applications. From a data-based analysis point of view, anomalies can be defined as data points that represent non-typical events, that is, abnormalities, with respect to the rest of the considered observations. The importance of anomaly detection relies on the fact that abnormal data highlights potentially undesirable situations in regard to the underlying physical phenomena under observation, which can have severe consequences for human beings, nature, infrastructures or information. This review article intends to provide a comprehensive overview of recent work on anomaly detection in a critical sector that is experiencing a deep digital transformation: the energy sector. With that, 52 articles have been reviewed, most of which focus on renewable energy generation, building energy consumption and energy storage. Interestingly, artificial intelligence-based approaches are found in ensemble schemes, where different models are combined for the maximization of the anomaly detection performance, oftentimes including deep learning (DL) models. However, under-represented trends and knowledge gaps are also identified, underscoring the lack of articles referring to specific energy application domains, such as critical infrastructures and electric vehicle (EV) charging infrastructure, and open issues for specific methodologies, such as explainability and applicability for deep learning anomaly detection solutions. Further, emerging concepts are highlighted and future research directions are identified.

Part of the book: Anomaly Detection