Résumé: Fault detection in rotating machinery is critical to reliability and safety. However, it faces difficulties due to complex, noisy fault signatures, non-stationary behavior, and the impracticality of obtaining large labeled datasets, limiting the effectiveness of both traditional and deep learning-based methods in real-world applications. This paper introduces a novel approach that combines Variational Mode Decomposition (VMD) and Long Short-Term Memory (LSTM) networks to improve gear and bearing defect detection, filling a gap in fault diagnostics by effectively handling limited training data. VMD decomposes signals into intrinsic mode functions (IMFs), while LSTM classifies fault types and severity levels based on time-domain features extracted from the IMFs. Tested on the Case Western Reserve University Dataset (CWRUDS) for bearing defects and the Laboratory of Mechanics and Structures Dataset (LMSDS) for combined gear and bearing defects, the method outperforms vibratory analysis and conventional classifiers such as MLP, 1D-CNN, 2D-CNN, and standalone LSTM. The results show that the VMD-LSTM model is superior at reliably detecting defects and accurately diagnosing faults in complex, data-limited scenarios, making it a promising solution for machinery health monitoring.

Résumé: In this study, we propose an advanced and recent method for processing no stationary and nonlinear signals in an industrial environment, based on cyclostationary analysis. The proposed technique is initially used to signals measured on defective bearings on a test bench. Finally, the industrial application realized on a valid production machine for large-scale, beyond that of the laboratory. The main objective of this work is to respond to the demand of a large FERTIAL industrial group where vibration analysis has been used to analyze the signal vibration measured on the journal bearings of a reducer GVAB420 of a turbo alternator GZ1164.

Résumé: This paper introduces an innovative spectral analysis control approach aimed at monitoring and diagnosing machine malfunctions to prevent potential failures. The research was conducted on a critical machine in a major industrial enterprise. The proposed method involves the use of a new indicator, called Overall Level (OL), that evaluates the machine’s condition before any operation. This study showcases practical methodologies for transitioning from time-based maintenance to predictive strategies, furnishing actionable insights into machine condition. This yields tangible advantages for the industry in terms of optimizing maintenance practices and enhancing asset productivity. Additionally, various methods, including vibration analysis, performance monitoring, and data analysis, are employed to identify the causes of issues and recommend solutions to enhance the reliability of the turbo compressor. The results provide a clear representation of the machine’s vibration state for diagnostic purposes. This noteworthy intervention underscores the potential of incorporating the measured and calculated values of the OL indicator across three specifically chosen frequency bands. To achieve this objective, the average value is employed as an indicator, contributing to the enhancement of reliability and longevity of critical industrial machinery. In this context, the novelty of the findings resides in the advanced diagnostic capabilities of the turbocompressor, thereby augmenting the efficacy of condition-based preventive maintenance for the BP 103 J turbine. The ultimate goal is to extend the equipment’s lifespan, improve the efficiency of the rotating machinery fleet, reduce maintenance costs, and enhance parameters such as availability and reliability through the support of an electronic maintenance system.

Résumé: Early detection and prompt intervention by maintenance engineers to mitigate the impact of breakdowns while enhancing overall operational efficiency remain critical challenges. This study proposes an innovative approach aiming at improving the diagnosis of gear faults. The objective is to assess the sensitivity and performance of traditional indicators in comparison to cyclostationarity, examining their impact on noise levels and vibrational signatures. The initial phase involves simulating gear signals under various conditions such as amplitude, rotation frequency, and meshing frequency, providing the foundation for a thorough analysis of indicator sensitivity and performance. In the second phase, both scalar and cyclostationary indicators were calculated. First, these indicators were compared against simulated signals, and second, their sensitivity and roughness were evaluated using signals measured on the bearings of 101 BJR reducers. This approach revealed that cyclostationary indicators are more sensitive than scalar indicators, suggesting an opportunity to improve the prediction of signal roughness throughout the production process. By introducing new possibilities to enhance the reliability of vibrational measurements, this method contributes to advancing the diagnosis of gear faults.

Résumé: The early detection of defects in rotating machines before their severity reaches an intolerable threshold has become the priority of the maintenance crew, hence the need to apply or even adapt new reliable methods that can meet this requirement. In this paper, a hybrid approach based on the complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN), the optimized wavelet multi-resolution analysis (OWMRA), and the Hilbert transform is proposed. The proposed approach is initially applied on experimental signals measured on faulty bearings of the Machinery Fault Simulator (MFS) test rig. Two types of defects are considered: on the outer race and on the inner race. An industrial application carried out on a production machine validates on a large scale, exceeding that of the laboratory, the proposed method. It was then possible to detect potential faults on a turbo-alternator group GZ 1164.2 operating in real industrial conditions in the largest fertilizer products company in Algeria. The obtained results show the power of the proposed approach to get reconstructed signals less noisy, containing more information and highlighting the precise nature of the defect and its severity in different configurations, regardless of its type. Finally, the main contribution of this study is to show that the application of the proposed approach is very practical on signals measured on real industrial rotating machines to detect the potential mechanical faults.

Résumé: the aim of this paper is to propose a comparative study between three advanced signal processing methods for the vibratory diagnosis of rotating machines working in industrial conditions. Cyclostationary analysis, empirical mode decomposition (EMD), and complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) are then applied for the detection of mechanical defects of a turbofan machine in the biggest fertilizer company in Algeria. These methods proved their efficiency for the diagnosis of specific defects, like rolling bearing and gear defects in laboratory test rigs, but their application in industrial field remains limited. The application of these methods on vibratory signals measured in low, medium, and high-frequency range allowed determining the efficiency of each method to diagnose the occurrence of different defects manifested in the three considered frequency ranges …

Résumé: The analysis of vibratory signals is an important subject in the field of researches on the diagnosis of mechanical breakdowns in rotating machinery. A number of signal processing methods are classified as classical tools, such as statistical analysis, FFT, envelope analysis (EA), and some time-frequency methods such as wavelet multi-resolution analysis (WMRA). These methods have shown their limitations when it is desired to analyze non-stationary and non-linear signals. In this paper, we introduce an advanced signal processing method, called cyclostationary analysis, which is probably one of the most recent tools used in vibratory diagnosis, especially in the case of non-stationary running conditions. A comparative study between five different methods and the cyclostationarity analysis is first conducted, choosing as criteria the advantages, disadvantages, and their contributions to the detection of mechanical defects. This comparative study is carried out on vibratory signals measured on defective rolling bearings kindly provided by the Bearing Data Center. In response to the request of a great industrial group, the cyclostationarity is used to analyze vibratory signals measured on a turbo-alternator working in real conditions. The experimental results obtained from this study confirm the ability of the proposed methodology to diagnose real mechanical defects in an industrial environment, in comparison with the results obtained by the application of the other classical methods mentioned previously.

Résumé: In this study, we present an application on the use of spectral analysis as aid to diagnosis and decision-making before a failure occurs due to a bad monitoring. The work is done on a strategic machine of a largest industrial company. This paper proposes the use of a new indicator that leads viewing the status of the machine before making an intervention and then exploiting the results as a handy reference. These values show the true state of the machine in terms of vibration diagnosis. The application of these decision-making references to a fan motor and its inclusion in the scorecard maintenance avoids unnecessary repairs caused by random decisions. The objective is to increase the life of the equipment and reduce the cost of production, resulting in the improvement of parameters such as the average level of functioning in an e-maintenance system. The comparison of the machine reliability before detecting the anomaly and the machine reliability after the proposal to use the new indicator and the intervention of the maintenance shows a best amelioration.

Résumé: The cyclostationarity method is used in this paper for the diagnosis of a turbo-alternator working in industrial environment for the detection of the defects generated by rolling bearings, journal bearings, and gears. This study shows the advantage of using such analysis as an aid to diagnosis and decision making before a failure caused by bad vibration monitoring of rotating machinery can be produced. In fact, a cyclostationary signal has some hidden periodicities, which mean that it is not strictly periodic, but some statistical properties of the signal are periodic. This periodicity identifies the spectral correlation by integrating the modulation intensity distribution function that depends only of the cyclic frequency, which is an indicator of the presence of modulations. The method was initially applied on a theoretical signal simulating a single bearing fault. The experimental validation is then performed on the …

Résumé: La surveillance et le diagnostic des défauts des machines tournantes appartiennent aux programmes de la maintenance conditionnelle, qui sont basés à 75% sur l’analyse des vibrations. Depuis plusieurs années les chercheurs travaillent sur l’amélioration ou le développement de nouvelles méthodes de traitement des signaux vibratoires. Le travail proposé dans cette thèse vise à établir un diagnostic fiable de deux installations industrielles, un turboventilateur et un turboalternateur, qui ont été la cause principale de plusieurs arrêts de la production et ont même poussé les responsables de maintenance à prendre la décision de changer des pièces coûteuses sans résoudre le problème. Nous proposons pour le diagnostic de la première installation un nouvel indicateur scalaire basé sur le calcul du niveau global moyen des vibrations mesurées dans les différents paliers et dans les trois directions principales. Ce nouvel indicateur a révélé l’existence d’un désalignement entre le moteur et le ventilateur.

Résumé: In this paper, an innovative approach is presented to enhance gear fault diagnosis using the cyclostationarity method. The first part of this study focuses on simulating gear signals under various conditions, allowing exploration of signal characteristics in vibration measurements. Spectral analyses and statistical calculations are performed to extract both classical and cyclostationary indicators. In the second part, the cyclostationarity method is applied to signals recorded at the gearbox bearings, clearly revealing the presence of faults. The results from these experiments demonstrate that cyclostationarity indicators can be leveraged to improve the prediction of signal roughness during the production process. This approach thus opens up new possibilities to enhance the reliability of vibration measurements and refine gear fault diagnosis.

Résumé: In this paper we have used a frequency modulation method for detecting faults in the plain bearings and the gear teeth defects. This method is based mainly on the analysis of some cyclostationarity non-stationary signals. Indeed, a cyclostationary signal has hidden periodicities; that is to say, it is not periodic in the strict sense but some statistical properties of the signal are periodic. This frequency is used to identify the spectral correlation which has the advantage of being a function of a single variable frequency instead of two. The experimental validation is performed on the basis of signals measured in an industrial environment (turbogenerator). The application of this method to non stationed signals has helped to highlight very clearly the presence of defects in the bearings of the gearbox, which has been difficult to demonstrate by spectral analysis.

Résumé:

Résumé:

Résumé:

Résumé:

Résumé:

Résumé:

Résumé:

Résumé:

Résumé:

Résumé: