Fault Diagnosis Precision Research Articles

Alpha‐Stable Distribution and Multifractal Detrended Fluctuation Analysis‐Based Fault Diagnosis Method Application for Axle Box Bearings

A railway vehicle’s key components, such as wheelset treads and axle box bearings, often suffer from fatigue failures. If these faults are not detected and dealt with in time, the running safety of the railway vehicle will be seriously affected. To detect these components’ early failure and extend their fatigue life, a regular maintenance becomes critical. Currently, the regular maintenance of axle box bearings mainly depends on manual off‐line inspection, which has low working efficiency and precision of fault diagnosis. In order to improve the maintenance efficiency and effectiveness of railway vehicles, this study proposes a method of integrating the vibration monitoring system of the axle box bearing in the underfloor wheelset lathe, where the integration scheme and work flow of the system are introduced followed by the detailed fault diagnosis method and application examples. Firstly, the band‐pass filter and envelope analysis is successively performed on the original signal acquired by an accelerometer. Secondly, the alpha‐stable distribution (ASD) and multifractal detrended fluctuation analysis (MFDFA) analysis of the envelope signal are performed, and five characteristic parameters with significant stability and sensitivity are extracted and then brought into the least squares support vectors machine based on particle swarm optimization to determine the state of the bearing quantitatively. Finally, the effectiveness of the method is validated by bench test data. The results demonstrated that the proposed method can accomplish effective diagnosis of axle box bearings’ fault location and fault degree and can yield better diagnosis accuracy than the single method of ASD or MFDFA.

Fault Diagnosis of Intershaft Bearings Using Fusion Information Exergy Distance Method

For the fault diagnosis of intershaft bearings, the fusion information exergy distance method (FIEDM) is proposed by fusing four information exergies, such as singular spectrum exergy, power spectrum exergy, wavelet energy spectrum exergy, and wavelet space spectrum exergy, which are extracted from acoustic emission (AE) signals under multiple rotational speeds and multimeasuring points. The theory of FIEDM is investigated based on four information exergy distances under multirotational speeds. As for rolling bearings, four faults and one normal condition are simulated on a birotor test rig to collect the AE signals, in which the four faults are inner ring fault, outer ring fault, rolling element fault, and inner race‐rolling element coupling fault. The faults of the intershaft bearings are analyzed and diagnosed by using the FIEDM. From the investigation, it is demonstrated that the faults of the intershaft bearings are accurately diagnosed and identified, and the FIEDM is effective for the analysis and diagnosis of intershaft bearing faults. Furthermore, the fault diagnosis precision of intershaft bearings becomes higher with increasing rotational speed.

Intelligent Fault Diagnosis Under Varying Working Conditions Based on Domain Adaptive Convolutional Neural Networks

Traditional intelligent fault diagnosis works well when the labeled training data (source domain) and unlabeled testing data (target domain) are drawn from the same distribution. However, in many real-world applications, the working conditions can vary between training and testing time. In this paper, we address the issues of intelligent fault diagnosis when the data at training and testing time do not come from the same distribution as a domain adaptation problem using domain adaptive convolutional neural networks (DACNN). Our proposed DACNN consists of three parts: a source feature extractor, a target feature extractor, and a label classifier. We adopt a two-stage training process to obtain strong fault-discriminative and domain-invariant capacity. First, we obtain fault-discriminative features by pre-training the source feature extractor with labeled source training examples to minimize the label classifier error. Then, in the domain adaptive fine-tuning stage, we train the target feature extractor to minimize the squared maximum mean discrepancy between the output of the source and target feature extractor, such that the instances sampled from the source and target domains have similar distributions after the mapping. Furthermore, to enable training efficiency in domain adaptation, the layers between the source and target feature extractors in our DACNN are partially untied during the training stage. Experiments on the bearing and gearbox fault data showed that DACNN can achieve high fault diagnosis precision and recall under different working conditions, outperforming other intelligent fault diagnosis methods. We also demonstrate the ability to visualize the learned features and the networks to better understand the reasons behind the remarkable performance of our proposed model.

Sensor fault diagnosis of aero-engine based on divided flight status

Fault diagnosis and safety analysis of an aero-engine have attracted more and more attention in modern society, whose safety directly affects the flight safety of an aircraft. In this paper, the problem concerning sensor fault diagnosis is investigated for an aero-engine during the whole flight process. Considering that the aero-engine is always working in different status through the whole flight process, a flight status division-based sensor fault diagnosis method is presented to improve fault diagnosis precision for the aero-engine. First, aero-engine status is partitioned according to normal sensor data during the whole flight process through the clustering algorithm. Based on that, a diagnosis model is built for each status using the principal component analysis algorithm. Finally, the sensors are monitored using the built diagnosis models by identifying the aero-engine status. The simulation result illustrates the effectiveness of the proposed method.

Bilevel Feature Extraction-Based Text Mining for Fault Diagnosis of Railway Systems

A vast amount of text data is recorded in the forms of repair verbatim in railway maintenance sectors. Efficient text mining of such maintenance data plays an important role in detecting anomalies and improving fault diagnosis efficiency. However, unstructured verbatim, high-dimensional data, and imbalanced fault class distribution pose challenges for feature selections and fault diagnosis. We propose a bilevel feature extraction-based text mining that integrates features extracted at both syntax and semantic levels with the aim to improve the fault classification performance. We first perform an improved $\chi^{2}$ statistics-based feature selection at the syntax level to overcome the learning difficulty caused by an imbalanced data set. Then, we perform a prior latent Dirichlet allocation-based feature selection at the semantic level to reduce the data set into a low-dimensional topic space. Finally, we fuse fault features derived from both syntax and semantic levels via serial fusion. The proposed method uses fault features at different levels and enhances the precision of fault diagnosis for all fault classes, particularly minority ones. Its performance has been validated by using a railway maintenance data set collected from 2008 to 2014 by a railway corporation. It outperforms traditional approaches.

A Novel Personalized Diagnosis Methodology Using Numerical Simulation and an Intelligent Method to Detect Faults in a Shaft

Personalized medicine is a hot topic to develop a medical procedure for healthcare. Motivated by molecular dynamics simulation-based personalized medicine, we propose a novel numerical simulation-based personalized diagnosis methodology and explain the fundamental procedures. As an example, a personalized fault diagnosis method is developed using the finite element method (FEM), wavelet packet transform (WPT) and support vector machine (SVM) to detect faults in a shaft. The shaft unbalance, misalignment, rub-impact and the combination of rub-impact and unbalance are investigated using the present method. The method includes three steps. In the first step, Theil’s inequality coefficient (TIC)-based FE model updating technique is employed to determine the boundary conditions, and the fault-induced FE model of the faulty shaft is constructed. Further, the vibration signals of the faulty shaft are obtained using numerical simulation. In the second step, WPT is employed to decompose the vibration signal into several signal components. Specific time-domain feature parameters of all of the signal components are calculated to generate the training samples to train the SVM. Finally, the measured vibration signal and its components decomposed by WPT serve as a test sample to the trained SVM. The fault types are finally determined. In the simulation of a simple shaft, the classification accuracy rates of unbalance, misalignment, rub-impact and the combination of rub-impact and unbalance are 93%, 95%, 89% and 91%, respectively, whereas in the experimental investigations, these decreased to 82%, 87%, 73% and 79%. In order to increase the fault diagnosis precision and general applicability, further works are continuously improving the personalized diagnosis methodology and the corresponding specific methods.

Fault diagnosis based on a novel weighted support vector data description with fuzzy adaptive threshold decision

The fault diagnosis of generator units is critical to guarantee the high efficiency of the electric system. However, detailed fault samples are difficult to obtain, and the distribution of fault samples usually shows the characteristics of unevenness and unbalance, which may lead to low fault diagnosis precision. Nevertheless, it has been seldom considered in the traditional classifier of fault diagnosis for generator units until now. In this paper, a novel fault classifier of weighted support vector data description (SVDD) with fuzzy adaptive threshold decision is proposed and applied in the fault diagnosis of generator units. To tackle the drawback that SVDD is sensitive to the distribution of samples, a novel SVDD model based on a complex weight is proposed. The complex weight is assigned with local density and size-based weight, while local density of each data point is obtained with the k-nearest neighbour approach and the size-based weight of each data point is computed according to the proportion of classes. Then the conventional SVDD is reformulated with the complex weights. Furthermore, new decision rules based on the relative distance and fuzzy adaptive threshold decision are applied to identify the class of testing samples. Finally, the proposed method is applied in the identification of several standard datasets, as well as the fault diagnosis for a turbo-generator unit. Experimental results and the engineering application reveal that the proposed method shows good performance in accuracy and universality, and is suitable for the fault diagnosis of generator units.

Diagnosis of Ship Generator Optimized Neural Network Based on Multi-population Chaos Genetic Algorithm

In view of fault diagnosis of the ship generator , the paper proposes improved fault diagnosis method of ship generator ,which is Optimized Neural Network based on Multipopulation Chaos Genetic Algorithm. The results prove that the method effectively solves low precision,slow constringency and local minimum of neural network and improves global search ability, optimizes the rate and precision of fault diagnosis. The method has a certain application prospect for the ship power system generator fault diagnosis.

Distributed Monitoring and Intelligent Fault Diagnosis System for Large-Scale Electro-Hydraulic Devices

In this paper, a distributed monitoring and intelligent diagnosis system is designed for large-scale electro-hydraulic devices, based on the utilization of information fusion technology. As a result of using multi-sensor data level, feature level and decision-making level information fusion methods, both precision and real-time performance of condition monitoring and fault diagnosis have been greatly improved, which leads to an advance in the accuracy of monitoring and diagnosis.

Hybrid computing techniques for fault detection and isolation, a review

The classical model-based methods had often proven to be unable to provide acceptable solutions to modern fault diagnosis systems. Therefore, model-free or soft computing techniques such as fuzzy logic, Artificial Neural Networks (ANNs) and Genetic Algorithms (GAs) had become more attractive in industrial applications of fault diagnosis. In this paper, three SC schemes are explored to solve the problem of detecting unprecedented changes and finding the failed state components. First, individual fuzzy systems, ANN and GA are implemented on a fault diagnosis scheme. Then, hybrids of these techniques are applied to enhance the fault diagnosis precision. This approach allows gaining critical information about fault presence or its absence in the shortest possible time. The proposed scheme was simulated and evaluated extensively on a benchmark laboratory scale coupled-two-tank system. The results are encouraging, showing especially, that hybrid GA+ANFIS (GANFIS), outperformed significantly the other techniques.

A Comparative Research on Condenser Fault Diagnosis Based on Three Different Algorithms

In view of artificial neural network, there are some deficiencies in condenser fault diagnosis. The BP neural network used for condenser fault diagnosis is highly nonlinear pattern recognition and high precision of fault diagnosis. The PSO-BP neural network can effectively solve the problem of BP neural network that training time is long and training process is easy to fall into the local minimum. The training results of PSO-BP network in convergence speed and convergence effect are significantly improved. Under the condition of small samples, the calculation results of SVM method are better than the calculation result of the other two methods. Although the recognition ratio of improved PSO-BP(2) and SVM is the same, training time of improved PSO-BP(2) is longer than training time of SVM. The generalization ability of SVM is stronger, and the efficiency of SVM is higher than the neural network. With MATLAB programming, three different algorithms, which are BP neural network, PSO-BP neural network and SVM, are studied and compared for the performance of condenser fault diagnosis. In the models of this study, the research results show that condenser fault diagnosis based on SVM has the fastest convergence speed and the best accuracy.

The Motor Rotor Fault Diagnosis Method Based on Linear Mixing Blind Separation Model

The detection precision of fault diagnosis based on frequency spectral analysis of stator current is easily restricted by noise jamming and frequency resolution. A fault diagnosis method for induction motor based on linear mixing model is proposed to resolve this problem. The fault characteristic signals are separated from the motor stator current by Fast-ICA algorithm and its amplitude is calculated according to the estimated mixing matrix. The fault diagnosis is achieved by difference of the amplitude on the normal state and the fault state of the motor. In this paper, the fault diagnosis of the broken rotor bars faults is used as an example to explain the conclusion as mentioned. Experiment result shows that the broken-rotor-bar fault can be diagnosed by the algorithm with better effect on the condition of short data block.

Quantitative Diagnosis of Rotor Vibration Fault Using Process Power Spectrum Entropy and Support Vector Machine Method

To improve the diagnosis capacity of rotor vibration fault in stochastic process, an effective fault diagnosis method (named Process Power Spectrum Entropy (PPSE) and Support Vector Machine (SVM) (PPSE-SVM, for short) method) was proposed. The fault diagnosis model of PPSE-SVM was established by fusing PPSE method and SVM theory. Based on the simulation experiment of rotor vibration fault, process data for four typical vibration faults (rotor imbalance, shaft misalignment, rotor-stator rubbing, and pedestal looseness) were collected under multipoint (multiple channels) and multispeed. By using PPSE method, the PPSE values of these data were extracted as fault feature vectors to establish the SVM model of rotor vibration fault diagnosis. From rotor vibration fault diagnosis, the results demonstrate that the proposed method possesses high precision, good learning ability, good generalization ability, and strong fault-tolerant ability (robustness) in four aspects of distinguishing fault types, fault severity, fault location, and noise immunity of rotor stochastic vibration. This paper presents a novel method (PPSE-SVM) for rotor vibration fault diagnosis and real-time vibration monitoring. The presented effort is promising to improve the fault diagnosis precision of rotating machinery like gas turbine.

Rolling Bearing Fault Diagnosis under Variable Conditions Using Hilbert-Huang Transform and Singular Value Decomposition

Fault diagnosis precision for rolling bearings under variable conditions has always been unsatisfactory. To solve this problem, a fault diagnosis method combining Hilbert-Huang transform (HHT), singular value decomposition (SVD), and Elman neural network is proposed in this paper. The method includes three steps. First, instantaneous amplitude matrices were obtained by using HHT from rolling bearing signals. Second, the singular value vector was acquired by applying SVD to the instantaneous amplitude matrices, thus reducing the dimension of the instantaneous amplitude matrix and obtaining the fault feature insensitive to working condition variation. Finally, an Elman neural network was applied to the rolling bearing fault diagnosis under variable working conditions according to the extracted feature vector. The experimental results show that the proposed method can effectively classify rolling bearing fault modes with high precision under different operating conditions. Moreover, the performance of the proposed HHT-SVD-Elman method has an advantage over that of EMD-SVD or WPT-PCA for feature extraction and Support Vector Machine (SVM) or Extreme Learning Machine (ELM) for classification.

Technology for Transformer Fault Diagnosis with Detection and Monitoring Data Fusion

Fault diagnosis is an effective means to assure the safe operation of power system. In this paper, a detection and monitoring of transformer fault diagnosis of multi-source data fusion technology is introduced. Based on the correlation distance, to calculate the relationship between detection and monitoring data and equipment failure, using the weight function to set fault indicators, status online monitoring data over the fault indicators, judging the equipment what kind of failure is happened, improving the precision of fault diagnosis.

Fault Diagnosis of Wind Turbine Gearbox Based on Least Square Support Vector Machine with Genetic Algorithm

Gearbox affect the normal operation of the wind turbines, to study the fault diagnosis, support vector method was used. Parameters selection is very important and decides the fault diagnosis precision. In order to overcome the blindness of man-made choice of the parameters in least squares support vector machine (LSSVM) and improve the accuracy and efficiency of fault diagnosis, a method based on LSSVM trained by genetic algorithm was proposed. This method searches the optimized parameters in LSSVM by taking advantage of the genetic algorithms powerful global searching ability. The research is provided using this method on the fault diagnosis of wind turbine gearbox and compared with the diagnostic method of LSSVM. The experimental results show that the method achieves a higher diagnostic accuracy.

Application of Fault Diagnosis for Air Blower Based on Genetic Fuzzy Neural Network

The way of fault characteristic parameters fuzzy processing and optimizing the weights and thresholds of ANN by GA are studied. As a result, the convergent rate and convergent precision are greatly increased. Application to the fault diagnosis of a air blower system shows the new model overcomes the low learning rate and local optima of BP algorithm, and the fault diagnosis precision is effectively improved.

Fault Diagnosis Method for Network Control Systems Applying SVM Disturbance Compensation

According to a network control system--NCS with short time-delay and packet loss, an state observer is designed firstly in this paper to obtain a state estimation error equation which is equivalent to an asynchronous dynamical system having event incidence constraint. Secondly, SVM is used to identify interferences of the NCS. Finally, making the identification result as compensation term adding to the state observer can make the residual only represent fault term under ideal condition and increase the robustness of NCS for interference, which can improve the fault diagnosis precision. The simulation results prove that the designed observer can diagnose faults effective and the disturbance compensation based on SVM has attained the expected effect.

Application of the CEMS on Fault Diagnosis for Rotary Machine

In order to monitor states of the rotary machine in time and ensure its performance, it is very important to analyze the evolution from the normal state to the fault state. In the paper, a new method is proposed to improve the precision of fault diagnosis. Firstly, the character extraction with mutative scales (CEMS) is applied to achieve the characteristic values. Secondly, the Hidden Semi-Markov model is built to identify the different running states. Thirdly, the new method is compared with the traditional one by the example of bushing abrasion of the connecting rod in diesel engine. According to the simulation and experiment researches, it indicates that the signal characters with more state information can be obtained pertinently by using the CEMS. And the accuracy of recognition is 97.33% in the 150 test samples, improved evidently than the traditional one. And the new character extraction method can be used in the technology domain widely.

Fault Diagnosis of Broken Bars in Induction Motors Using EMD

This paper presents a diagnosis method based on EMD, in order to improve the precision of fault diagnosis in induction motors. After the Park vector analogue signal of stator current is decomposed by EMD, 2sf fault characteristic value is abstracted accurately. Simulation results show that this proposed method is of high sensitivity, efficiently solves the problem that frequencies of fault and fundamental wave are close and difficult to be decomposed.