Dimensionless features and comprehensive fuzzy-based models for fault diagnosis of rolling element bearings under varying operating conditions.

A review on signal processing techniques utilized in the fault diagnosis of rolling element bearings

Vibration analysis is an established method for fault detection and diagnosis of rolling element bearings. However, it is an expert oriented exercise. To relieve the experts, the use of Artificial Intelligence (AI) techniques such as deep neural networks, especially convolutional neural networks (CNN) have gained the attention of researchers because of their image classification and recognition capability. Most researchers convert the vibration signal into representative time frequency vibration images such as spectrograms and scalograms. These images are used as inputs to train the CNN model for fault diagnosis. Commonly, fault diagnosis is performed under same operating conditions, where models are trained and deployed for prediction under the same operating conditions. However, outside the laboratory environment, in real world applications, different operating conditions, such as variable speed, may be encountered. With the change in speed, the characteristic frequencies of the vibration signal will also change, which will result in changing the vibration image. Consequently, the performance of the CNN model may drop significantly for prediction under different operating conditions. Accessing the training data from all potential operating conditions may not be feasible for most real-world applications. Therefore, there is a need to find some signal properties which are invariant to change in operating conditions and only change due to change in health state so that models trained under one set of operating conditions may predict correctly under different operating conditions. This paper proposes a defect diagnosis method for rolling element bearings, under variable operating conditions (speed and load) based on CNN and order maps. These maps exhibit consistent properties under varying speed; therefore, they can be used to train the CNN model for fault diagnosis under variable speed. The effect of load change on these order maps is experimentally studied and it is found that the proposed method can undertake fault diagnosis on rolling element bearings under variable speeds and loads with high accuracy.

Condition monitoring and fault diagnosis play an important role in the health management of mechanical equipment. However, the robust performance of data-driven-based methods with unknown fault inputs remains to be further improved. In this paper, a novel approach of condition monitoring and fault diagnosis is proposed for rolling element bearings based on an improved ensemble empirical mode decomposition (IEEMD), which is able to solve the non-intrinsic mode function problem of EEMD. In this method, IEEMD is applied to process the primordial vibration signals collected from rolling element bearings at first. Then the correlation analysis and data fusion technology are introduced to extract statistical features from these decomposition results of IEEMD. Finally, a complete self-zero space model is constructed for the condition monitoring and fault diagnosis of rolling element bearings. Experiments are implemented on a mechanical fault simulator to demonstrate the reliability and effectiveness of the proposed method. The experimental results show that the proposed method can not only diagnose known faults but also monitor unknown faults with strong robust performance.

Condition monitoring and fault diagnosis of rolling element bearings are very important to ensure proper working of different types of machinery. Condition monitoring of rotating machines is mainly based on the analysis of machine vibration. The vibration signals from the mechanical fault generally comprise periodic impulses with specified characteristic frequency corresponds to a particular defect. But due to heavy noise in the industry, the vibration signals have a very low signal-to-noise ratio. Hence, it requires an appropriate technique to extract the impulses from the noisy signal. This article emphasized on the fault diagnosis of rolling element bearings having some specific size of defects on various bearing elements using the complex Morlet wavelet analysis. The phase and amplitude map of the complex Morlet wavelet are utilized for identification and diagnosis of the fault in the rolling element bearing. The amplitude and phase map corresponding to the complex Morlet wavelet are found to show unique informative signatures in the presence of bearing faults. The classification technique based on artificial neural network and support vector machine for rolling element bearing fault detection is presented in this article. The classification results of bearing faults clearly indicate that support vector machine has a more precise bearing fault classification ability than artificial neural network.

Fault diagnosis of rolling element bearing using cyclic autocorrelation and wavelet transform

Machine condition monitoring in remote locations and harsh environment where network infrastructure is not feasible, or hardwired network connectivity is not possible, wireless communications provides an alternative which also offer installation cost savings, improve reliability and quicker deployment. This paper describes the implementation of wireless sensor network (WSN) for early fault diagnosis of rolling element bearings based on signal autocorrelation technique. A low-power 2.4 GHz wireless HART transceiver, a low-cost wireless vibration transmitter, 26.76 mv/g accelerometers and a 1420 wireless gateway with AMS software was implemented. The research describes the methodology of acquiring peak values data in high frequency region. The noise was averaged out by applying four-time averaging and natural frequencies or fault frequencies of bearing elements was captured. The experimental results show that the signal autocorrelation algorithm can successfully diagnose the roller bearing faults at early stage on wireless signal transfer. As the raw data was processed before wireless transmission on analyzing unit and spectrum was transferred in JPG format on display unit, minimum power consumption has been noted. The technique provided a better alternative of wired system for real time condition monitoring of roller bearings in rotating equipment installed in remote area.

Rolling element bearing is one of the most important and common components in rotary machines, whose failures can cause both personal damage and economic loss. This paper focuses on condition monitoring and fault diagnosis of rolling element bearing in order to detect the failure ahead of time and estimate the fault location accurately when failure occurs. Wavelet energy entropy is introduced into the field of mechanical condition monitoring and SOM network is used in fault diagnosis of rolling element bearing. In order to validate the effectiveness of the proposed method, a bearing accelerated life test is performed on the accelerated bearing life tester(ABLT-1A). The results indicate that wavelet energy entropy has better performance and can forecast fault development earlier compared with kurtosis and RMS of the vibration signal, while SOM network, which has a advantage of visualization, can distinguish bearing fault type well.

Ensemble empirical mode decomposition (EEMD) represents a valuable aid in empirical mode decomposition (EMD) and has been widely used in fault diagnosis of rolling element bearings. However, the intrinsic mode functions (IMFs) generated by EEMD often contain residual noise. In addition, adding different white Gaussian noise to the signal to be analyzed probably produces a different number of IMFs, and different number of IMFs makes difficult the averaging. To alleviate these two drawbacks, complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) was previously presented. Utilizing the advantages of CEEMDAN in extracting weak characteristics from noisy signals, a new fault diagnosis method of rolling element bearings based on CEEMDAN is proposed. With this method, a particular noise is added at each stage and after each IMF extraction, a unique residue is computed. In this way, this method solves the problem of the final averaging and obtains IMFs with less noise. A simulated signal is used to illustrate the effectiveness of the proposed method, and the decomposition results show that the method obtains more accurate IMFs than the EEMD. To further demonstrate the proposed method, it is applied to fault diagnosis of locomotive rolling element bearings. The diagnosis results prove that the method based on CEEMDAN may reveal the fault characteristic information of rolling element bearings better.

Nonlinear characteristics are ubiquitous in the vibration signals produced by rolling element bearings. Fractal dimensions are effective tools to illustrate nonlinearity. This paper proposes a new approach based on Multiscale General Fractal Dimensions (MGFDs) to realize fault diagnosis of rolling element bearings, which are robust to the effects of variation in operating conditions. The vibration signals of bearing are analyzed to extract the general fractal dimensions in multiscales, which are in turn utilized to construct a feature space to identify fault pattern. Finally, bearing faults are revealed by pattern recognition. Case studies are carried out to evaluate the validity and accuracy of the approach. It is verified that this approach is effective for fault diagnosis of rolling element bearings under various operating conditions via experiment and data analysis.

PurposeThe purpose of this article is to present a new application of pursuit‐based analysis for diagnosing rolling element bearing faults.Design/methodology/approachIntelligent diagnosis of rolling element bearing faults in rotating machinery involves the procedure of feature extraction using modern signal processing techniques and artificial intelligence technique‐based fault detection and identification. This paper presents a comparative study of both the basis and matching pursuits when applied to fault diagnosis of rolling element bearings using vibration analysis.FindingsFault features were extracted from vibration acceleration signals and subsequently fed to a feed forward neural network (FFNN) for classification. The classification rate and mean square error (MSE) were calculated to evaluate the performance of the intelligent diagnostic procedure. Results from the basis pursuit fault diagnosis procedure were compared with the classification result of a matching pursuit feature‐based diagnostic procedure. The comparison clearly illustrates that basis pursuit feature‐based fault diagnosis is significantly more accurate than matching pursuit feature‐based fault diagnosis in detecting these faults.Practical implicationsIntelligent diagnosis can reduce the reliance on experienced personnel to make expert judgements on the state of the integrity of machines. The proposed method has the potential to be extensively applied in various industrial scenarios, although this application concerned rolling element bearings only. The principles of the application are directly translatable to other parts of complex machinery.Originality/valueThis work presents a novel intelligent diagnosis strategy using pursuit features and feed forward neural networks. The value of the work is to ease the burden of making decisions on the integrity of plant through a manual program in condition monitoring and diagnostics particularly of complex pieces of plant.

The rolling element bearing is a core component of many systems such as aircraft, train, steamboat, and machine tool, and their failure can lead to reduced capability, downtime, and even catastrophic breakdowns. Due to misoperation, manufacturing deficiencies, or the lack of monitoring and maintenance, it is often found to be the most unreliable component within these systems. Therefore, effective and efficient fault diagnosis of rolling element bearings has an important role in ensuring the continued safe and reliable operation of their host systems. This study presents a trace ratio criterion-based kernel discriminant analysis (TR-KDA) for fault diagnosis of rolling element bearings. The binary immune genetic algorithm (BIGA) is employed to solve the trace ratio problem in TR-KDA. The numerical results obtained using extensive simulation indicate that the proposed TR-KDA using BIGA (called TR-KDA-BIGA) can effectively and efficiently classify different classes of rolling element bearing data, while also providing the capability of real-time visualization that is very useful for the practitioners to monitor the health status of rolling element bearings. Empirical comparisons show that the proposed TR-KDA-BIGA performs better than existing methods in classifying different classes of rolling element bearing data. The proposed TR-KDA-BIGA may be a promising tool for fault diagnosis of rolling element bearings.

Condition monitoring and intelligent diagnosis of rolling element bearings under constant/variable load and speed conditions

Condition monitoring and fault diagnosis of rolling element bearings (REBs) are at present very important to ensure the reliability of rotating machinery. This paper presents a new pattern classification approach for bearings diagnostics, which combines Mathematical Morphology (MM) and Multi-output Adaptive Neuro Fuzzy Inference System (M-ANFIS) classifier. MM is used for filtering Vibration signals, which acquired through the accelerometers mounted on the bearing housing. In this regard, to have an effective morphological operator, the structure elements (SEs) are selected based on the Kurtosis value. Then, to design an automated fault diagnosis structure, the features of this filtered signal, are extracted and used in the M-ANFIS model to learn and classify the bearing condition. The MM method overcomes the drawbacks of other signal processing methods and the M-ANFIS model can handle variation conditions. The experimental results indicate that the proposed strategy not only reduces the error rate but also is robust to changes of load, speed and size of defects.

Fault diagnosis of rolling element bearings (REBs), one type of essential mechanical element, has been actively researched; recent research has focused on the use of deep-learning-based approaches. However, conventional deep-learning-based fault-diagnosis approaches are vulnerable to various operating speeds, which greatly affect the vibration characteristics of the system studied. To solve this problem, previous deep-learning-based studies have usually been carried out by increasing the complexity of the model or diversifying the task of the model. Still, limitations remain because the reason of increasing complexity is unclear and the roles of multiple tasks are not well-defined. Therefore, this study proposes a multi-head de-noising autoencoder-based multi-task model for robust diagnosis of REBs under various speed conditions. The proposed model employs a multi-head de-noising autoencoder and multi-task learning strategy to robustly extract features under various speed conditions, while effectively disentangling the speed- and fault-related information. In this research, we evaluate the proposed method using the signals measured from bearing experiments under various speed conditions. The results of the evaluation study show that the proposed method outperformed conventional methods, especially when the training and test datasets have large discrepancies in their operating conditions.

Sparsity guided empirical wavelet transform for fault diagnosis of rolling element bearings