Scale-Fractal Detrended Fluctuation Analysis for Fault Diagnosis of a Centrifugal Pump and a Reciprocating Compressor.

Reciprocating compressors and centrifugal pumps are rotating machines used in industry, where fault detection is crucial for avoiding unnecessary and costly downtime. A novel method for fault classification in reciprocating compressors and multi-stage centrifugal pumps is proposed. In the feature extraction stage, raw vibration signals are processed using multi-fractal detrended fluctuation analysis (MFDFA) to extract features indicative of different types of faults. Such MFDFA features enable the training of machine learning models for classifying faults. Several classical machine learning models and a deep learning model corresponding to the convolutional neural network (CNN) are compared with respect to their classification accuracy. The cross-validation results show that all models are highly accurate for classifying the 13 types of faults in the centrifugal pump, the 17 valve faults, and the 13 multi-faults in the reciprocating compressor. The random forest subspace discriminant (RFSD) and the CNN model achieved the best results using MFDFA features calculated with quadratic approximations. The proposed method is a promising approach for fault classification in reciprocating compressors and multi-stage centrifugal pumps.

Modelos de variables latentes en patrones de alimentación y actividad física en niños/adolescentes: una revisión sistemática / Models of latent variables used for eating and physical activity patterns in children/ adolescents: a systematic review

Introducción. Debido a la poca evidencia sobre el modelamiento de los patrones de alimentación y actividad física (AF), basado en variables latentes, el presente estudio de revisión pretende describir las técnicas estadísticas aplicadas para modelar estos patrones en niños y adolescentes y valorar su calidad metodológica. Materiales y métodos. La búsqueda se realizó en bases de datos electrónicas (Science Direct, PubMed, SCOPUS, Web of science y Cochrane) con las palabras "diet", 'physical activity', children y 'latent variable'. Se incluyeron artículos que utilizaron modelos estadísticos basados en variables latentes para analizar patrones de alimentación y AF en niños y adolescentes sanos, publicados entre 2014­2019, en inglés o español. Resultados. Entre los 27 artículos seleccionados, el Modelo de Ecuaciones Estructurales (MEE) fue el más utilizado (77,78%); seguido del Modelo de Perfil Latente (7,41%), mientras, el restante, 14,81% aplican el Modelo del Factor Común, Modelo Ecológico y el Modelo de Regresión Logística Multinivel. El MEE fue aplicado a 12 de los 16 artículos con enfoque de AF, y en 7 de los 9 artículos con enfoque de Alimentación. El 48,15% de los estudios sí justificaba el uso del modelo, y el 37,04% poseen una calidad "Excelente" (cumplen el 85% o más de los ítems de STROBE). Conclusiones. El MEE fue el más utilizado para abstraer los patrones de AF y alimentación en niños y adolescentes, sin embargo, solo la mitad de los artículos justifica su pertinencia. Las guías de reporte de estudios deberían evaluar la calidad metodológica de los modelos estadísticos aplicados(AU)

Introduction. Due to the limited evidence on the modeling of eating and physical activity (PA) patterns based on latent variables, the present review study aims to describe the statistical techniques applied to model these patterns in children and adolescents and to assess their methodological quality. Materials and methods. The search was performed in electronic databases (Science Direct, PubMed, SCOPUS, Web of science and Cochrane) with the words 'diet', 'physical activity', children and 'latent variable'. We included articles that used statistical models based on latent variables to analyze diet and PA patterns in healthy children and adolescents, published between 2014-2019, in English or Spanish. Results. Among the 27 selected articles, the Structural Equation Model (SEM) was the most used (77.78%); followed by the Latent Profile Model (7.41%), while, the remaining 14.81% applied the Common Factor Model, Ecological Model and Multilevel Logistic Regression Model. The SEM was applied to 12 of the 16 articles with PA approach, and in 7 of the 9 articles with eating approach. The 48.15% of studies did justify the use of the model, and 37.04% were classified as "Excellent" quality (meet 85% or more of the STROBE items). Conclusions. The SEM was the most commonly used to model the PA and eating patterns in children and adolescents, however, only half of the articles justify their relevance. Study reporting guidelines should evaluate the methodological quality of the statistical models applied(AU)

From fault detection to one-class severity discrimination of 3D printers with one-class support vector machine.

The lack of faulty condition data reduces the feasibility of supervised learning for fault detection or fault severity discrimination in new manufacturing technologies. To deal with this issue, one-class learning arises for building binary discriminative models using only healthy condition data. However, these models have not been extrapolated to severity discrimination. This paper proposes to extend OCSVM, which is typically used for fault detection, to 3D printer fault severity discrimination. First, a set of features is extracted from a set of normal signals. An optimized OCSVM model is obtained by tuning the kernel and model hyperparameters. The resulting models are evaluated for fault detection and fault severity discrimination using a proposed performance evaluation approach. Experimental comparisons for belt-based faults in 3D printers show that the distance to the hyperplane has the information to discriminate the severity level, and its use is feasible. The proposed hyperparameter optimization technique improves the OCSVM for fault detection and severity discrimination compared to some other methods.

Using a Support Vector Machine Based Decision Stage to Improve the Fault Diagnosis on Gearboxes.

Gearboxes are mechanical devices that play an essential role in several applications, e.g., the transmission of automotive vehicles. Their malfunctioning may result in economic losses and accidents, among others. The rise of powerful graphical processing units spreads the use of deep learning-based solutions to many problems, which includes the fault diagnosis on gearboxes. Those solutions usually require a significant amount of data, high computational power, and a long training process. The training of deep learning-based systems may not be feasible when GPUs are not available. This paper proposes a solution to reduce the training time of deep learning-based fault diagnosis systems without compromising their accuracy. The solution is based on the use of a decision stage to interpret all the probability outputs of a classifier whose output layer has the softmax activation function. Two classification algorithms were applied to perform the decision. We have reduced the training time by almost 80% without compromising the average accuracy of the fault diagnosis system.

Fault Diagnosis for Rotating Machinery Using Vibration Measurement Deep Statistical Feature Learning.

Fault diagnosis is important for the maintenance of rotating machinery. The detection of faults and fault patterns is a challenging part of machinery fault diagnosis. To tackle this problem, a model for deep statistical feature learning from vibration measurements of rotating machinery is presented in this paper. Vibration sensor signals collected from rotating mechanical systems are represented in the time, frequency, and time-frequency domains, each of which is then used to produce a statistical feature set. For learning statistical features, real-value Gaussian-Bernoulli restricted Boltzmann machines (GRBMs) are stacked to develop a Gaussian-Bernoulli deep Boltzmann machine (GDBM). The suggested approach is applied as a deep statistical feature learning tool for both gearbox and bearing systems. The fault classification performances in experiments using this approach are 95.17% for the gearbox, and 91.75% for the bearing system. The proposed approach is compared to such standard methods as a support vector machine, GRBM and a combination model. In experiments, the best fault classification rate was detected using the proposed model. The results show that deep learning with statistical feature extraction has an essential improvement potential for diagnosing rotating machinery faults.

Multi-Stage Feature Selection by Using Genetic Algorithms for Fault Diagnosis in Gearboxes Based on Vibration Signal.

There are growing demands for condition-based monitoring of gearboxes, and techniques to improve the reliability, effectiveness and accuracy for fault diagnosis are considered valuable contributions. Feature selection is still an important aspect in machine learning-based diagnosis in order to reach good performance in the diagnosis system. The main aim of this research is to propose a multi-stage feature selection mechanism for selecting the best set of condition parameters on the time, frequency and time-frequency domains, which are extracted from vibration signals for fault diagnosis purposes in gearboxes. The selection is based on genetic algorithms, proposing in each stage a new subset of the best features regarding the classifier performance in a supervised environment. The selected features are augmented at each stage and used as input for a neural network classifier in the next step, while a new subset of feature candidates is treated by the selection process. As a result, the inherent exploration and exploitation of the genetic algorithms for finding the best solutions of the selection problem are locally focused. The Sensors 2015, 15 23904 approach is tested on a dataset from a real test bed with several fault classes under different running conditions of load and velocity. The model performance for diagnosis is over 98%.