Due to the real working conditions, the collected mechanical fault datasets are actually limited and always highly imbalanced, which restricts the diagnosis accuracy and stability. To solve these problems, we present an imbalanced fault diagnosis method based on the generative model of conditional-deep convolutional generative adversarial network (C-DCGAN) and provide a study in detail. Deep convolutional generative adversarial network (DCGAN), based on traditional generative adversarial networks (GAN), introduces convolutional neural network into the training for unsupervised learning to improve the effect of generative networks. Conditional generative adversarial network (CGAN) is a conditional model obtained through introducing conditional extension into GAN. C-DCGAN is a combination of DCGAN and CGAN. In C-DCGAN, based on the feature extraction ability of convolutional networks, through the structural optimization, conditional auxiliary generative samples are used as augmented data and applied in machine fault diagnosis. Two datasets (Bearing dataset and Planetary gear box dataset) are carried out to validate. The simulation experiments showed that the improved performance is mainly due to the generated signals from C-DCGAN to balance the dataset. The proposed method can deal with imbalanced fault classification problem much more effectively. This model could improve the accuracy of fault diagnosis and the generalization ability of the classifier in the case of small samples and display better fault diagnosis performance.

A case study of conditional deep convolutional generative adversarial networks in machine fault diagnosis

Safety and reliability are absolutely important for modern sophisticated systems and technologies. Therefore, malfunction monitoring capabilities are instilled in the system for detection of the incipient faults and anticipation of their impact on the future behavior of the system using fault diagnosis techniques. In particular, state-of-the-art applications rely on the quick and efficient treatment of malfunctions within the equipment/system, resulting in increased production and reduced downtimes. This paper presents developments within Fault Detection and Diagnosis (FDD) methods and reviews of research work in this area. The review presents both traditional model-based and relatively new signal processing-based FDD approaches, with a special consideration paid to artificial intelligence-based FDD methods. Typical steps involved in the design and development of automatic FDD system, including system knowledge representation, data-acquisition and signal processing, fault classification, and maintenance related decision actions, are systematically presented to outline the present status of FDD. Future research trends, challenges and prospective solutions are also highlighted.

A review on fault detection and diagnosis techniques: basics and beyond

The demand of condition monitoring of induction motors (IM) is progressively increasing to maintain the performance of several important sectors in industry. This issue is of great importance since it prevents IM from failing and breaking down. As most of IM faults occur in bearings, the bearing fault detection (BFD) has become the main topic targeting the optimization of unscheduled downtime and maintenance cost of IM. Besides, emphasizing the causes and predicting failure consequences depend on the identification of the fault type. This paper is motivated by the advances in signal processing techniques and machine-learning systems. This study proposes a novel hybrid approach for BFD based on Optimized Stationary Wavelet Packet Transform for feature extraction and artificial immune system nested within support vectors machines for fault classification. The motor current signatures analysis offers a cost-effective method for BFD. To evaluate the approach, the current signals were collected under various bearing conditions and load levels. The experiment results prove the efficiency of the proposed approach.

Distinct Bearing Faults Detection in Induction Motor by a Hybrid Optimized SWPT and aiNet-DAG SVM

This paper proposes a novel approach for detection of stator short-circuit faults in three-phase induction motors. The method is based on two stages: feature extraction and classification by intelligent systems. First, mutual information is estimated from delayed stator current signals, which are used as inputs of C4.5 decision trees and multilayer perceptron neural networks in the second step. Several offline and online experimental tests are presented considering voltage unbalance, load torque variations, and 1% to 10% short-circuit levels. The obtained results corroborate the effectiveness of this new diagnostic approach.

Stator Short-Circuit Diagnosis in Induction Motors Using Mutual Information and Intelligent Systems

Enhanced generative adversarial networks for fault diagnosis of rotating machinery with imbalanced data

This paper describes a new diagnosis approach, the Root-Multiple Signal Classification (MUSIC) (RM) method, for identification of the progressive cracking in the bearing of induction motors. This approach has several advantages compared with the stator current spectral analysis using the conventional Periodogram method. Indeed, the main advantage of this approach is its very good frequency resolution for a very short acquisition time, something impossible to achieve with the conventional method. However, in order to reduce the computation time, which is the main drawback of the RM method, this method will be applied to only a specified frequency band; one that carries information about the sought fault. Experimental results show the effectiveness of the RM method on the reliability of the incipient bearing fault detection.

Induction Motor Bearing Fault Analysis Using a Root-MUSIC Method

Induction motor diagnosis using the Power Spectral Density (PSD) estimation based on the Fourier Transform calculation has been widely used as an analysis method for its simplicity and low computation time. However, the use of PSD is not recommended for processing non stationary signals (case of variable speed applications) and therefore the analysis with PSD is not reliable. To overcome this handicap, the Short Time Fourier Transform (STFT) is proposed in this paper; giving additional information on changes of the frequencies over time for stator current signal analysis. Furthermore, the use of a new approach called Maxima’s Location Algorithm is also proposed. This later will be associated with the STFT analysis to show only those harmonics with useful information on existing faults. This approach will be used in the diagnosis of bearing faults of a PWM inverter-fed induction motor operating at variable speed. Several experimental results in the transient state are carried out firstly to validate the results and secondly to illustrate the merits and effectiveness of the combined STFT/MLA proposed approach.

Bearing Fault Diagnosis of a PWM Inverter Fed-Induction Motor Using an Improved Short Time Fourier Transform

To show the reliability of new methods developed for the induction motors rotor faults diagnosis, a comparison is often made with the classical method based on the estimation of the Power Spectral Density by Periodogram (PSDP). The authors of the new methods always criticize the PSDP method for its drawbacks related to the smoothing and side effects of the selected window function. This paper aims at showing that the PSDP based on the Kaiser window associated with an adequate calculation of the shape parameter can overcome these drawbacks. This calculation which takes into account the characteristics of the stator current spectrum, will improve the detection of the faults while retaining the main advantage of the method, namely a fast calculation time. The experimental results obtained show the positive contribution of the proposed approach on the reliability of the diagnosis of the incipient faults.

Effect of Kaiser Window Shape Parameter for the Enhancement of Rotor Faults Diagnosis

Le diagnostic des defauts affectant les roulements a bille utilises dans les entrainements electriques est un axe de recherche tres etudie vu l’importance de ces organes dans le transfert d’energie. A cet effet, plusieurs techniques de diagnostic ont ete developpees. Parmi ces techniques on cite l’analyse vibratoire, approche tres utilisee en industrie, et l’analyse du courant statorique. Le but de cet article est d’effectuer une simple etude comparative sur les capacites et les performances de chaque technique dans le diagnostic des defauts de roulement a billes d’un moteur asynchrone. Les resultats experimentaux obtenus montrent que ces deux techniques sont complementaires vu que leur fiabilite depend de l’element defectueux constituant le roulement.

Comparaison entre la Technique Vibratoire et la Technique des Courants Statoriques : Application au Diagnostic des Roulements à Billes

Statistically, the outer race of the rolling-element bearing is the most sensitive element used in induction motors. Among the known diagnosis methods, the analysis of the stator current envelope based on its demodulation, makes it possible to directly obtain the signature of the searched faults without the fundamental effect. This paper discusses one of these methods known as the Wavelet-Kurtogram. For this purpose, a comparison between the Wavelet-Kurtogram and the Power Spectral Density using the Periodogram is carried out for bearing faults diagnosis. The experimental results obtained show the superiority and the effectiveness of Wavelet-Kurtogram in the detection of the outer race fault.

Mohammed-El-Amine Khodja

Papers

Induction Motor Bearing Fault Analysis Using a Root-MUSIC Method

Bearing Fault Diagnosis of a PWM Inverter Fed-Induction Motor Using an Improved Short Time Fourier Transform

Effect of Kaiser Window Shape Parameter for the Enhancement of Rotor Faults Diagnosis

Comparaison entre la Technique Vibratoire et la Technique des Courants Statoriques : Application au Diagnostic des Roulements à Billes

Outer Race Fault Diagnosis by Comparison between the Power Spectral Density and the Kurtogram