Fault Diagnosis Of Mechanical Systems Research Articles

Composite-Graph-Based Sparse Subspace Clustering for Machine Fault Diagnosis

Sparse subspace clustering (SSC) is an effective method to cluster sensing signal for fault diagnosis in mechanical systems. SSC is based on a global expression strategy describing each data point by other data points from all the potential clusters. A drawback of this strategy is that it generates nonzero elements in the nondiagonal blocks of the similarity matrix, thereby reducing the performance of the matrix for data discrimination. To improve SSC's performance, a composite-graph-based SSC (CG-SSC) method is developed by introducing distance among the data points into SSC, where the L1-norm of the sparse coefficients is replaced by the product of the data distance and the sparse coefficients. The distance serves as a constraint to the amplitude of the sparse coefficients. A sparse coefficient with small amplitude is assigned if the distance between the expressed data and the expressing data is large, and vice versa. This method supports the construction of a compact cluster. The effectiveness of the presented method is experimentally verified by data measured on several bearings and gearboxes with different types of faults. The method is also compared with classical clustering methods, and the results indicate its advantage for data sorting and clustering.

An experimental platform for statistical fault diagnosis in propelled mechanical systems: an academic exercise

Statistical data analysis for fault diagnosis in mechanical systems is a fundamental tool, for instance, in applied mechanical engineering. In order to capture a feasible data set, a well designed electronic instrumentation and excitation system signal stages are mandatory. Hence, one objective of this paper is to develop a low cost vibration sensor based on an inductive LC-tank oscillator (a resonant inductive-capacitive electronic circuit carefully designed to produce an harmonic electrical signal), and then to tune an effective excitation system signal to our experimental platform. This platform uses a propelled drone motor mounted on a beam structure to emulate a propelled rotating machine. Essentially, two data set were acquired. One for the healthy behaviour of the developed system, and the other for a programmed faulty scenario. This defective case was realized by introducing a small mechanical fault in one blade extreme of the mechanical propelled system. To note, this faulty scenario is almost impossible to deduce by just seen the raw data. The other objective of this paper is to analyze the obtained data sets by utilizing a statistical data analysis tool. Then, by employing box-plot diagrams, the healthy and faulty cases become evidenced. Finally, and due to we are proposing a low-cost academic experimental platform for fault diagnosis based on data analysis, our platform’s toll was around 120 euros. Hence, this platform results applicable to teach data analysis from dynamical systems.

Learning transferable features in deep convolutional neural networks for diagnosing unseen machine conditions

Recent years have witnessed increasing popularity and development of deep learning spanning through various fields. Deep networks, and in particular convolutional neural network (CNN) have also achieved many state-of-the-art competition results in the intelligent fault diagnosis of mechanical systems. However, most of the existing studies have been performed with the assumption that the same distribution holds for both the training data and the test data, which is not in accord with situations in real diagnosis tasks. To tackle this problem, a transfer learning framework based on pre-trained CNN, which leverages the knowledge learned from the training data to facilitate diagnosing a new but similar task, is presented in this work. First, the CNN is trained on large datasets to learn the hierarchical features from the raw data. Then, the architecture and weights of the pre-trained CNN are transferred to new tasks with proper fine-tuning instead of training a network from scratch. To adapt the pre-trained CNN in a specific case, three transfer learning strategies are discussed and compared to investigate the applicability as well as the significance of feature transferability from the different levels of a deep structure. The case studies show that the proposed framework can transfer the features of the pre-trained CNN to boost the diagnosis performance on unseen machine conditions in terms of diverse working conditions and fault types.

A convolutional neural network based feature learning and fault diagnosis method for the condition monitoring of gearbox

Feature extraction plays a vital role in intelligent fault diagnosis of mechanical system. Nevertheless, traditional feature extraction methods suffer from three problems, which are (1) the requirements of domain expertise and prior knowledge, (2) the sensitive to the changes of mechanical system and (3) the limitations of mining new features. It is attractive and meaningful to investigate an automatic feature extraction method, which can adaptively learn features from raw data and discover new fault-sensitive features. Deep learning has been widely used in image analysis and speech recognition with great success. The key advantage of this method lies into the ability of mining representative information and sensitive features from raw data. However, the application of deep learning in feature leaning for mechanical diagnosis is still few, and limited studies have been carried out to compare the effectiveness of feature leaning with various data types. This paper will focus on developing a convolutional neural network (CNN) to learn features directly from frequency data of vibration signals and testing the different performance of feature learning from raw data, frequency spectrum and combined time-frequency data. Manual features from time domain, frequency domain and wavelet domain as well as three common intelligent methods are used as comparisons. The effectiveness of the proposed method is validated through PHM 2009 gearbox challenge data and a planetary gearbox test rig. The results demonstrate that the proposed method is able to learn features adaptively from frequency data and achieve higher diagnosis accuracy than other comparative methods.

An approach to robust fault diagnosis in mechanical systems using computational intelligence

In this paper a novel approach to design robust fault diagnosis systems in mechanical systems using historical data and computational intelligence techniques is presented. First, the pre-processing of the data to remove the outliers is performed with the aim of reducing the classification errors. To accomplish this objective, the Density Oriented Fuzzy C-Means (DOFCM) algorithm is used. Later on, the Kernel Fuzzy C-Means (KFCM) algorithm is used to achieve greater separability among the classes, and reducing the classification errors. Finally, an optimization process of the parameters used in the training state by the DOFCM and KFCM for improving the classification results is developed using the bioinspired algorithm Ant Colony Optimization. The proposal was validated using the DAMADICS (Development and Application of Methods for Actuator Diagnosis in Industrial Control Systems) benchmark. The satisfactory results obtained indicate the feasibility of the proposal.

Analysis of the feasibility of an array of MEMS microphones to machinery condition monitoring or fault diagnosis

During the last decades, vibration analysis has been used to evaluate condition monitoring and fault diagnosis of complex mechanical systems. The problem associated with these analysis methods is that the employed sensors must be in contact with the vibrant surfaces. To avoid this problem, the current trend is the analysis of the noise, or the acoustic signals, which are directly related with the vibrations, to evaluate condition monitoring and/or fault diagnosis of mechanical systems. Both, acoustic and vibration signals, obtained from a system can reveal information related with its operation conditions. Using arrays formed by digital MEMS microphones, which employ acquisition/processing systems based on FPGA, allows creating systems with a high number of sensors paying a reduced cost. This work studies the feasibility of the use of acoustic images, obtained by an array with 64 MEMS microphones (8x8) in a hemianechoic chamber, to detect, characterize, and, eventually, identify failure conditions in machinery. The resolution obtained to spatially identify the problem origin in the machine under test. The acoustic images are processed to extract different feature patterns to identify and classify machinery failures.

Rolling Element Bearing Fault Diagnosis Using Improved Manifold Learning

Fault feature can be extracted by traditional manifold learning algorithms, which construct neighborhood graphs by Euclidean distance (ED). It is difficult to get an excellent dimensionality reduction result when processed data has strong correlations. In order to improve the effect of dimensionality reduction and increase accuracy of bearing fault diagnosis in mechanical systems, an improved manifold learning method based on Mahalanobis distance (MD) is proposed. In this paper, we use time-domain analysis and frequency-domain analysis to construct high-dimensional feature vectors in the first step. Then, MD is used to replace ED in neighborhood construction of manifold learning. After using the improved manifold learning method, low-dimensional feature vectors can be extracted. Finally, fault diagnosis of rolling element bearing can be made by applying the K-nearest neighbor classifier. In part of experiment, to verify the efficiency of the improved manifold learning methods, artificial data sets and rolling element bearing fault data are adopted. The experimental comparison results of the improved manifold learning algorithm and the traditional algorithm prove that the proposed method is more effective in rolling element bearing fault diagnosis.

Online sequential prediction of bearings imbalanced fault diagnosis by extreme learning machine

Diagnosis of bearings generally plays an important role in fault diagnosis of mechanical system, and machine learning has been a promising tool in this field. In many real applications of bearings fault diagnosis, the data tend to be online imbalanced, which means, the number of fault data is much less than the normal data while they are all collected in online sequential way. Suffering from this problem, many traditional diagnosis methods will get low accuracy of fault data which acts as the minority class in the collected bearing data. To address this problem, an online sequential prediction method for imbalanced fault diagnosis problem is proposed based on extreme learning machine. This method introduces the principal curve and granulation division to simulate the flow distribution and overall distribution characteristics of fault data, respectively. Then a confident over-sampling and under-sampling process is proposed to establish the initial offline diagnosis model. In online stage, the obtained granules and principal curves are rebuilt on the bearing data which are arrived in sequence, and after the over-sampling and under-sampling process, the balanced sample set is formed to update the diagnosis model dynamically. A theoretical analysis is provided and proves that, even existing information loss, the proposed method has lower bound of the model reliability. Simulation experiments are conducted on IMS bearing data and CWRU bearing data. The comparative results demonstrate that the proposed method can improve the fault diagnosis accuracy with better effectiveness and robustness than other algorithms.

Time-frequency atoms-driven support vector machine method for bearings incipient fault diagnosis

Bearing plays an essential role in the performance of mechanical system and fault diagnosis of mechanical system is inseparably related to the diagnosis of the bearings. However, it is a challenge to detect weak fault from the complex and non-stationary vibration signals with a large amount of noise, especially at the early stage. To improve the anti-noise ability and detect incipient fault, a novel fault detection method based on a short-time matching method and Support Vector Machine (SVM) is proposed. In this paper, the mechanism of roller bearing is discussed and the impact time frequency dictionary is constructed targeting the multi-component characteristics and fault feature of roller bearing fault vibration signals. Then, a short-time matching method is described and the simulation results show the excellent feature extraction effects in extremely low signal-to-noise ratio (SNR). After extracting the most relevance atoms as features, SVM was trained for fault recognition. Finally, the practical bearing experiments indicate that the proposed method is more effective and efficient than the traditional methods in weak impact signal oscillatory characters extraction and incipient fault diagnosis.

Partly Duffing Oscillator Stochastic Resonance Method and Its Application on Mechanical Fault Diagnosis

Due to the fact that the slight fault signals in early failure of mechanical system are usually submerged in heavy background noise, it is unfeasible to extract the weak fault feature via the traditional vibration analysis. Stochastic resonance (SR), as a method of utilizing noise to amplify weak signals in nonlinear dynamical systems, can detect weak signals overwhelmed in the noise. However, based on the analysis of the impact of noise intensity on SR effect, it is concluded that the detection results are dramatically limited by the noise intensity of measured signals, especially for incipient fault feature of mechanical system with poor working environment. Therefore, this paper proposes a partly Duffing oscillator SR method to extract the fault feature of mechanical system. In this method, to locate the appearance of weak fault feature and decrease noise intensity, the permutation entropy index is constructed to select the measured signals for the input of Duffing oscillator system. Then, according to the regulation of system parameters, a reasonable match between the selected signals and Duffing oscillator model is achieved to produce a SR phenomenon and realize the fault diagnosis of mechanical system. Experiment results demonstrate that the proposed method achieves a better effect on the fault diagnosis of mechanical system.

Time—window Complexity and its Application in the Fault Diagnosis of Bearing

Different from conventional spectral method, complexity analysis treaded with the signals’ time domain structure feature. Before, due to the no stationary and the unevenness of state space of mechanical signals, the state information is likely to get a loss. Now, the time-window complexity is proposed to overcome certain limitations of complexity itself in some extent. It will help to extract the state features of mechanical systems in different states. The concept and algorithm of time-window complexity is introduced detailed. The way of applying the time-window complexity for fault diagnosis is discussed. The mechanical signals of ball bearing with slight flaw and that of four kinds of impact-rubbing states of a typical rotor are then studied. The results show that time-window complexity can reflect the early fault of ball bearing, and differentiate the four kinds of rotor impact-rubbing states properly, which provides another effective way for fault diagnosis of mechanical systems.

A Novel Clustering Method Combining ART with Yu’s Norm for Fault Diagnosis of Bearings

Clustering methods have been widely applied to the fault diagnosis of mechanical system, but the characteristic that the number of cluster needs to be determined in advance limits the application range of the method. In this paper, a novel clustering method combining the adaptive resonance theory (ART) with the similarity measure based on the Yu’s norm is presented and applied to the fault diagnosis of rolling element bearings, which can be adaptive to generate the number of cluster by the vigilance parameter test. Time-domain features, frequency-domain features, and time series model parameters are extracted to demonstrate the fault-related information about the bearings, and then considering the irrelevance or redundancy of some features many salient features are selected by an improved distance discriminant technique and input into the proposed clustering method to diagnose the faults of bearings. The experiment results confirmed that the proposed clustering method can diagnose the fault categories accurately and has better diagnosis performance compared with fuzzy ART and Self-Organizing Feature Map (SOFM).

ON-LINE PERFORMANCE ASSESSMENT AND FAULT DIAGNOSIS OF MECHANICAL SYSTEMS

This study integrates sensors, signal capture equipment, industrial computers and machinery health check-up software to develop an On-line Performance Assessment and Fault Diagnosis of Mechanical System, helping engineers predict mechanical conditions. Physical quantities captured by the sensors is utilized to process physical signals, and the Wavelet Packet Energy method is used for the feature extraction of non-stationary signals in coordination with the Principal Component Analysis for feature selection. This study establishes On-line Performance Assessment and Fault Diagnosis of Mechanical System based on Discriminant Analysis which is able to immediately determine the mechanical performance. When abnormal mechanical conditions occur, Bayesian Network will be activated to construct error diagnostic model and determine possible causes of error or malfunction of the machinery. Finally, the system is applied to the fan motor, high-speed spindle motor and AC motor of the machine tool. Experimental results show that the theory can effectively diagnose mechanical performance remarkable with an accuracy rate of 92.50% or higher.

Actuator fault diagnosis in mechanical systems — Fault power estimation approach

A systematic method is introduced to deal with actuator faults in a class of nonlinear mechanical control systems. The presence of the fault is detected by monitoring the fault induced modifications in the control system’s energy balance. From the energy deviation the residual generator estimates the power of the fault. For fault isolation and identification the set of fault power pattern signals is defined. Using signal processing methods the elements of this set are compared to the residual signal to determine the characteristics and the magnitude of the fault. The method does not require assumptions on the system’s model and it can handle parameter uncertainties. The applicability of the proposed method is shown through mechatronic examples.

De-Noising and Analyzing for Non-Stationary Signal Based on MATLAB

This paper studies the application of wavelet analysis in the fault diagnosis of mechanical system which describes the principle of wavelet analysis and its application in fault diagnosis of gear mechanism. It can identify and eliminate the failure by analyzing the vibration signal obtained from the fault simulation experiment of gear transmission system. Wavelet analysis in MATLAB can extract some important fault characteristics that the other methods cannot extract them. Application of wavelet analysis in the fault diagnosis of gear transmission system is effective by pretreating the characteristic information that extracted from gear transmission system.

Variable cosine windowing of intrinsic mode functions: Application to gear fault diagnosis

Gear pair is used for speed reduction or increasing torque and/or to change the direction of rotation. Gears are considered critical element in various mechanical systems. When gears are in use, the multi component vibration signals are generated. These vibration signals can be captured by mounting accelerometers at suitable locations. Vibration signal analysis is very effective tool in finding gear fault at early stage. The methods based on empirical mode decomposition (EMD) have been used for gear fault diagnosis in mechanical systems. The EMD method decomposes an original signal into different frequency-bands in time domain, known as intrinsic mode functions (IMFs). A serious problem in application of EMD is boundary distortion of IMFs. While doing statistical analysis of IMFs, boundary distortion may provide high values of statistical indicator (e.g. kurtosis, Sr, Sα), even if fault is not present. Several extension-based methods are employed to eliminate the boundary distortion problem. Extension-based methods cannot completely eliminate the boundary distortion, especially when the low-frequency component of the analyzed signal is weak. Recently, cosine window-based method has been proposed by which the boundary distortion can be controlled in boundaries of the signal and the middle component of it can be exactly decomposed into IMFs. The cosine window-based method works only for a particular IMF depending on the size of window. Since, in EMD process, the boundary distortion of successive IMFs increases, a variable cosine window is proposed in this paper to address the increasing boundary distortion problem. In the proposed method boundary distortion problem is minimized by using variable cosine window for all IMFs. The simulation and experimental results for three statistical indicators viz. kurtosis, Sr, Sα show that the proposed method based on variable cosine window is a powerful and reliable technique for fault diagnosis.

Research on Internet-Based Reconfigurable Distance Mechanical System Fault Diagnosis

In order to meet requirements of increasingly high-speed, large and intelligent mechanical equipments on fault diagnosis, the Internet-based reconfigurable mechanical system fault diagnosis program was presented. The overall structure and networking schema of distance mechanical fault diagnosis system were analyzed, and the distance fault diagnosis network model based on J2EE framework was also described. The structural model and reconfigurable manner of the reconfigurable distance diagnosis system was provided, which used CORBA component technology to achieve reconfiguration. The detail steps of system that take some type of diesel engine as diagnosis object was described, and the intelligent diagnosing methods were also researched. The Internet-based fault diagnosis technology effectively improves the efficiency and accuracy of diagnostic systems.

Application of Fault Phenomenon Vector Distance Discriminance in Mechanical System Fault Diagnosis

Aiming at the problem of diagnosis difficulty caused by too many factors of mechanical system, a kind of diagnosing method based on fault phenomenon was presented. The research on mechanical system fault phenomenon space arrived at conclusion that the emergency of each fault phenomenon subject to 0-1 distribution. Therefore, phenomenon vector corresponding to each fault formed cluster whose accumulation point is expectation of vector. After exclusion of abnormal vectors, the distance discrimination was used to fault diagnosis to establish expert system based on fault phenomenon vector. The confirmed result was return back to fault database so that the system achieve self-learning of real-time diagnosis experiences. Finally, the example on X-type hydraulic excavator proves that the diagnostic method has characteristics of good real-time, simple operation and high diagnostic accuracy.

Fault diagnostic system for rotating machine based on Wavelet packet transform and Elman neural network

An efficient fault diagnosis system is needed for industry because it can optimize the resources management and improve the performance of the system. In this study, a fault diagnostic system is proposed for rotating machine using wavelet packet transform (WPT) and elman neural network (ENN) techniques. In most fault diagnosis for mechanical systems, WPT is a well-known signal processing technique for fault detection and identification. In previous work, WPT can improve the continuous wavelet transform (CWT) used over a longer computing time and huge operand. It can also solve the frequency-band disagreement by discrete wavelet transform (DWT) only breaking up the approximation version. In the experimental work, the extracted features from the WPT are used as inputs in an Elman neural network. The results show that the scheme can reliably diagnose four different conditions and can be considered as an improvement of previous works in this field.

An intelligent fault diagnosis method based on wavelet packer analysis and hybrid support vector machines

In this paper, a new intelligent method for the fault diagnosis of the rotating machinery is proposed based on wavelet packet analysis (WPA) and hybrid support machine (hybrid SVM). In fault diagnosis for mechanical systems, information about stability and mutability can be further acquired through WPA from original signal. The faulty vibration signals obtained from a rotating machinery are decomposed by WPA via Dmeyer wavelet. A new multi-class fault diagnosis algorithm based on 1-v-r SVM approach is proposed and applied to rotating machinery. The extracted features are applied to hybrid SVM for estimating fault type. Compared to conventional back-propagation network (BPN), the superiority of the hybrid SVM method is shown in the success of fault diagnosis. The test results of hybrid SVM demonstrate that the applying of energy criterion to vibration signals after WPA is a very powerful and reliable method and hence estimating fault type on rotating machinery accurately and quickly.