Miguel Delgado-Prieto

Multiple-Fault Detection Methodology Based on Vibration and Current Analysis Applied to Bearings in Induction Motors and Gearboxes on the Kinematic Chain

Gearboxes and induction motors are important components in industrial applications and their monitoring condition is critical in the industrial sector so as to reduce costs and maintenance downtimes. There are several techniques associated with the fault diagnosis in rotating machinery; however, vibration and stator currents analysis are commonly used due to their proven reliability. Indeed, vibration and current analysis provide fault condition information by means of the fault-related spectral component identification. This work presents a methodology based on vibration and current analysis for the diagnosis of wear in a gearbox and the detection of bearing defect in an induction motor both linked to the same kinematic chain; besides, the location of the fault-related components for analysis is supported by the corresponding theoretical models. The theoretical models are based on calculation of characteristic gearbox and bearings fault frequencies, in order to locate the spectral components of the faults. In this work, the influence of vibrations over the system is observed by performing motor current signal analysis to detect the presence of faults. The obtained results show the feasibility of detecting multiple faults in a kinematic chain, making the proposed methodology suitable to be used in the application of industrial machinery diagnosis.

Multifault Diagnosis Method Applied to an Electric Machine Based on High-Dimensional Feature Reduction

Condition monitoring schemes are essential for increasing the reliability and ensuring the equipment efficiency in industrial processes. The feature extraction and dimensionality reduction are useful preprocessing steps to obtain high performance in condition monitoring schemes. To address this issue, this work presents a novel diagnosis methodology based on high-dimensional feature reduction applied to detect multiple faults in an induction motor linked to a kinematic chain. The proposed methodology involves a hybrid feature reduction that ensures a good processing of the acquired vibration signals. The method is performed sequentially. First, signal decomposition is carried out by means of empirical mode decomposition. Second, statistical-time-based features are estimated from the resulting decompositions. Third, a feature optimization is performed to preserve the data variance by a genetic algorithm in conjunction with the principal component analysis. Fourth, a feature selection is done by means of Fisher score analysis. Fifth, a feature extraction is performed through linear discriminant analysis. And, finally, sixth, the different considered faults are diagnosed by a Neural Network-based classifier. The performance and the effectiveness of the proposed diagnosis methodology is validated experimentally and compared with classical feature reduction strategies, making the proposed methodology suitable for industry applications.

Enhanced Industrial Machinery Condition Monitoring Methodology Based on Novelty Detection and Multi-Modal Analysis

This paper presents a condition-based monitoring methodology based on novelty detection applied to industrial machinery. The proposed approach includes both the classical classification of multiple a priori known scenarios, and the innovative detection capability of new operating modes not previously available. The development of condition-based monitoring methodologies considering the isolation capabilities of unexpected scenarios represents, nowadays, a trending topic able to answer the demanding requirements of the future industrial processes monitoring systems. First, the method is based on the temporal segmentation of the available physical magnitudes, and the estimation of a set of time-based statistical features. Then, a double feature reduction stage based on principal component analysis and linear discriminant analysis is applied in order to optimize the classification and novelty detection performances. The posterior combination of a feed-forward neural network and one-class support vector machine allows the proper interpretation of known and unknown operating conditions. The effectiveness of this novel condition monitoring scheme has been verified by experimental results obtained from an automotive industry machine.

Enhanced load forecasting methodology by means of probabilistic prediction intervals estimation

The improvement of the forecasting accuracy for prediction of future loads has been object of exhaustive study in the recent years, to the point that a wide variety of methodologies which have been proved to be valid and practical exists. However, most methodologies for demand forecasting do not handle uncertainties of the resulting model, which leads to a nonproper interpretation of the forecasted outcomes. In this context, this work presents a novel load forecasting methodology in order to quantify the model uncertainties and complement the resulting information by means of adaptive confidence intervals. First, an input selection technique based on Genetic Algorithms is used to select the best combination of inputs in order to obtain a state-of-the-art model by means of Adaptive Neuro-Fuzzy Inference Systems. Then the data space is analyzed in terms of error probability of the model outcomes. The principal component analysis is used to visualize the error probability in a 2-D map. Finally, an Artificial Neural Network is used to perform the identification of the error probability associated to new measurements. In conjunction with the forecasting model, the proposed classifier extends the resulting information with an adaptive confidence intervals and its probability distribution. The effectiveness of this enhanced load forecasting methodology has been verified by experimental data obtained from an automotive plant.

Vibration-Based Adaptive Novelty Detection Method for Monitoring Faults in a Kinematic Chain

This paper presents an adaptive novelty detection methodology applied to a kinematic chain for the monitoring of faults. The proposed approach has the premise that only information of the healthy operation of the machine is initially available and fault scenarios will eventually develop. This approach aims to cover some of the challenges presented when condition monitoring is applied under a continuous learning framework. The structure of the method is divided into two recursive stages: first, an offline stage for initialization and retraining of the feature reduction and novelty detection modules and, second, an online monitoring stage to continuously assess the condition of the machine. Contrary to classical static feature reduction approaches, the proposed method reformulates the features by employing first a Laplacian Score ranking and then the Fisher Score ranking for retraining. The proposed methodology is validated experimentally by monitoring the vibration measurements of a kinematic chain driven by an induction motor. Two faults are induced in the motor to validate the method performance to detect anomalies and adapt the feature reduction and novelty detection modules to the new information. The obtained results show the advantages of employing an adaptive approach for novelty detection and feature reduction making the proposed method suitable for industrial machinery diagnosis applications.

Time series forecasting by means of SOM aided Fuzzy Inference Systems

The forecast of industrial process time series represents a critical factor in order to assure a proper operation of the whole manufacturing chain, as it allows to act against any process deviation before it affects the final manufactured product. In this paper, in order to take advantage from process relations and improve forecasting performance, a prediction method based in Adaptive Neuro Fuzzy Inference System (ANFIS) and Self-Organizing Maps is presented. The novelties of the proposed method are based on considering, as an input of an ANFIS model, the interrelations of process variables regarding the signal that wants to be forecasted, by means of topology preservation SOM. An experimental study performed with real industrial data from a cooper manufacturing plant indicated the suitability of the proposed method in time series forecasting applications.

Diagnosis methodology for identifying gearbox wear based on statistical time feature reduction

Strategies for condition monitoring are relevant to improve the operation safety and to ensure the efficiency of all the equipment used in industrial applications. The feature selection and feature extraction are suitable processing stages considered in many condition monitoring schemes to obtain high performance. Aiming to address this issue, this work proposes a new diagnosis methodology based on a multi-stage feature reduction approach for identifying different levels of uniform wear in a gearbox. The proposed multi-stage feature reduction approach involves a feature selection and a feature extraction ensuring the proper application of a high-performance signal processing over a set of acquired measurements of vibration. The methodology is performed successively; first, the acquired vibration signals are characterized by calculating a set of statistical time-based features. Second, a feature selection is done by performing an analysis of the Fisher score. Third, a feature extraction is realized by means of the linear discriminant analysis technique. Finally, fourth, the diagnosis of the considered faults is done by means of a fuzzy-based classifier. The effectiveness and performance of the proposed diagnosis methodology are evaluated by considering a complete data set of experimental test, making the proposed methodology suitable to be applied in industrial applications with power transmission systems.

Diagnosis methodology based on statistical-time features and linear discriminant analysis applied to induction motors

The development of condition monitoring strategies is necessary to ensure the efficiency and reliability of the operation on electric machines. The feature calculation is an important signal processing step used to obtain a characterization related to the working condition of machinery. In order to address this issue, this work proposes a diagnosis methodology based on the calculation of a statistical-time set of features applied to identify the appearance of different faults in an induction motor. In the proposed methodology three acquired stator current signals are characterized by calculating its statistical-time features. Then, such statistical-time sets of features are compressed and represented into a 2-dimentional space through Linear Discriminant Analysis. And, finally a Neuro Fuzzy-based classifier is used to diagnose the different considered conditions. The performance of the proposed diagnosis methodology is evaluated in an experimental test bench; the obtained results make the proposed methodology suitable to be applied in industrial processes.The development of condition monitoring strategies is necessary to ensure the efficiency and reliability of the operation on electric machines. The feature calculation is an important signal processing step used to obtain a characterization related to the working condition of machinery. In order to address this issue, this work proposes a diagnosis methodology based on the calculation of a statistical-time set of features applied to identify the appearance of different faults in an induction motor. In the proposed methodology three acquired stator current signals are characterized by calculating its statistical-time features. Then, such statistical-time sets of features are compressed and represented into a 2-dimentional space through Linear Discriminant Analysis. And, finally a Neuro Fuzzy- based classifier is used to diagnose the different considered conditions. The performance of the proposed diagnosis methodology is evaluated in an experimental test bench; the obtained results make the proposed methodology suitable to be applied in industrial processes.