• Smart Structures and Systems
• /
• v.29 no.6
• /
• pp.757-766
• /
• 2022

To train deep learning algorithms, a sufficient number of data are required. However, in most engineering systems, the acquisition of fault data is difficult or sometimes not feasible, while normal data are secured. The dearth of data is one of the major challenges to developing deep learning models, and fault diagnosis in particular cannot be made in the absence of fault data. With this context, this paper proposes an anomaly detection methodology for rotating machines using only normal data with self-labeling. Since only normal data are used for anomaly detection, a self-labeling method is used to generate a new labeled dataset. The overall procedure includes the following three steps: (1) transformation of normal data to self-labeled data based on a pretext task, (2) training the convolutional neural networks (CNN), and (3) anomaly detection using defined anomaly score based on the softmax output of the trained CNN. The softmax value of the abnormal sample shows different behavior from the normal softmax values. To verify the proposed method, four case studies were conducted, on the Case Western Reserve University (CWRU) bearing dataset, IEEE PHM 2012 data challenge dataset, PHMAP 2021 data challenge dataset, and laboratory bearing testbed; and the results were compared to those of existing machine learning and deep learning methods. The results showed that the proposed algorithm could detect faults in the bearing testbed and compressor with over 99.7% accuracy. In particular, it was possible to detect not only bearing faults but also structural faults such as unbalance and belt looseness with very high accuracy. Compared with the existing GAN, the autoencoder-based anomaly detection algorithm, the proposed method showed high anomaly detection performance.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.32 no.11
• /
• pp.905-910
• /
• 2008

Vibration analysis is widely used in machinery diagnosis and the wavelet transform has also been implemented in many applications in the condition monitoring of machinery. In contrast to previous applications, this paper examines whether acoustic signal can be used effectively along vibration signal to detect the various local fault, in local fault of gearboxes using the wavelet transform. Moreover, envelop analysis is well known as useful tool for the detection of rolling element bearing fault. In this paper, a acoustic emission (AE) sensor is employed to detect gearbox damage by installing them around bearing housing at driven-end side. Signal processing is conducted by wavelet transform and enveloping to detect her fault all at once gearbox using AE signal.

• Journal of KSNVE
• /
• v.10 no.6
• /
• pp.985-990
• /
• 2000

The signals that can be obtained from rotating machines often convey the information of machine. For example, if the machine under investigation has faults, then these signals often have pulse signals, embedded in noise. Therefore the ability to detect the fault signal in noise is major concern of fault diagnosis of rotating machine, In this paper, minimum variance cepstrum (MV cepstrum) . which can easily detect impulse in noise, has been applied to detect the type of faults of ball bearing system. To test the performance of this technique. various experiments have been performed for ball bearing elements that have man made faults. Results show that minimum variance cepstrum can easily detect the periodicity due to faults and also shows the pattern of excitation by the faults.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.16 no.4 s.109
• /
• pp.404-413
• /
• 2006

Induction motors are the workhorse of our industry because of their versatility and robustness. The diagnosis of mechanical load and power transmission system failures is usually carried out through mechanical signals such as vibration signatures, acoustic emissions, motor speed envelope. The motor faults including mechanical rotor imbalances, broken rotor bar, bearing failure and eccentricities problems are reflected in electric, electromagnetic and mechanical quantities. The recent research has been directed toward electrical monitoring of the motor with emphasis on inspecting the stator current of the motor, The stator current spectrum has been widely used for fault detection in induction motor systems. The motor current signature analysis is the useful technique to assess machine electrical condition. This paper describes the motor condition detected by the current signatures Paralleled with vibration signatures analysis of induction motors with the roller bearing and the journal bearing type for large vertical pumps in power plant as examples to discuss for motor fault detection and diagnosis.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.19 no.2
• /
• pp.72-77
• /
• 2011

In this paper, we propose effective diagnosis algorithm for hub bearing fault in driving vehicle using acceleration signal and wheel speed signal measured in hub bearing unit or knuckle. This algorithm consists of differential, envelope and power spectrum method. We developed diagnosis system for realizing proposed algorithm. This system consists of input device including acceleration sensor and wheel speed sensor, calculation device using Digital Signal Processor (DSP) and display device using Personal Digital Assistant (PDA). Using this diagnosis system, a driver can see hub bearing fault(flaking) from the vibration in driving vehicle. With early repairing, he can keep good ride feeling and prevent accident of vehicle resulting from hub bearing fault.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2008.11a
• /
• pp.84-88
• /
• 2008

The roller bearings play an important role not only sustain radial or axial load of system, but carry out a rotatory movement as a various operating conditions. They happen that incipient faults which were caused by excessive load, manufacturing or assembling process's errors and many other reasons are created. The bearing faults make noise and vibration by a continuous collision of rotatory components, which can lower the quality and stability of auto-transmission. Therefore, it is important to detect the early fault as soon as possible. This paper presents a detecting method for the improvement in quality by developing the program which can be used to analyze and predict the vibrational characteristics caused by roller bearing faults. We completed development of the inspection system of vibration by appling the most efficient detecting methods and verified the system's reliability through experiments.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.19 no.3
• /
• pp.268-273
• /
• 2009

The roller bearings play an important role not only sustain radial or axial load of system, but carry out a rotatory movement as a various operating conditions. They happen that incipient faults which were caused by excessive load, manufacturing or assembling process's errors and many other reasons are created. The bearing faults make noise and vibration by a continuous collision of rotatory components, which can lower the quality and stability of auto-transmission. Therefore, it is important to detect the early fault as soon as possible. This paper presents a detecting method for the improvement in quality by developing the program which can be used to analyze and predict the vibrational characteristics caused by roller bearing faults. We completed development of the inspection system of vibration by applying the most efficient detecting methods and verified the system's reliability through experiments.

• Journal of Power Electronics
• /
• v.8 no.1
• /
• pp.101-107
• /
• 2008

In this paper, the features extracted from vibration time signals are used to detect the bearing fault condition. The decision tree is applied to diagnose the bearing status, which has the benefits of being an expert system that is based on knowledge history and is simple to understand. This paper also suggests a genetic algorithm (GA) as a method to reduce the number of features. In order to show the potentials of this method in both aspects of accuracy and simplicity, the reduced-feature decision tree is compared with the non reduced-feature decision tree and the PCA-based decision tree.

• Smart Structures and Systems
• /
• v.14 no.3
• /
• pp.377-395
• /
• 2014

Vibration-based fault detection and condition monitoring of rotating machinery, using statistical process control (SPC) combined with statistical pattern recognition methodology, has been widely investigated by many researchers. In particular, the discrete wavelet transform (DWT) is considered as a powerful tool for feature extraction in detecting fault on rotating machinery. Although DWT significantly reduces the dimensionality of the data, the number of retained wavelet features can still be significantly large. Then, the use of standard multivariate SPC techniques is not advised, because the sample covariance matrix is likely to be singular, so that the common multivariate statistics cannot be calculated. Even though many feature-based SPC methods have been introduced to tackle this deficiency, most methods require a parametric distributional assumption that restricts their feasibility to specific problems of process control, and thus limit their application. This study proposes a nonparametric multivariate control chart method, based on multiscale wavelet scalogram (MWS) features, that overcomes the limitation posed by the parametric assumption in existing SPC methods. The presented approach takes advantage of multi-resolution analysis using DWT, and obtains MWS features with significantly low dimensionality. We calculate Hotelling's $T^2$-type monitoring statistic using MWS, which has enough damage-discrimination ability. A bootstrap approach is used to determine the upper control limit of the monitoring statistic, without any distributional assumption. Numerical simulations demonstrate the performance of the proposed control charting method, under various damage-level scenarios for a bearing system.

• Journal of the Korean Society for Nondestructive Testing
• /
• v.32 no.1
• /
• pp.7-11
• /
• 2012

For fault detection about the abrasion stage of rotational machineries under the dynamic loading conditions unlike the traditional diagnosis method used in the past decade, the infrared thermographic method with its distinctive advantages in non-contact, non-destructive, and visible aspects is proposed. In this paper, by applying a rotating deep-grooved ball bearing, passive thermographic experiments were conducted as an alternative way to proceeding the traditional fault monitoring on spectrum analyzer. As results, the thermographic experiment was compared with the traditional vibration spectrum analysis to evaluate the efficiency of the proposed method. Based on the results obtained as NDT, the temperature characteristics and abnormal fault detections of the ball bearing according to the abrasion stage were analyzed.