Current Fault Diagnosis Research Articles

An Integrated Deep Learning Method towards Fault Diagnosis of Hydraulic Axial Piston Pump.

A hydraulic axial piston pump is the essential component of a hydraulic transmission system and plays a key role in modern industry. Considering varying working conditions and the implicity of frequent faults, it is difficult to accurately monitor the machinery faults in the actual operating process by using current fault diagnosis methods. Hence, it is urgent and significant to investigate effective and precise fault diagnosis approaches for pumps. Owing to the advantages of intelligent fault diagnosis methods in big data processing, methods based on deep learning have accomplished admirable performance for fault diagnosis of rotating machinery. The prevailing convolutional neural network (CNN) displays desirable automatic learning ability. Therefore, an integrated intelligent fault diagnosis method is proposed based on CNN and continuous wavelet transform (CWT), combining the feature extraction and classification. Firstly, CWT is used to convert the raw vibration signals into time-frequency representations and achieve the extraction of image features. Secondly, a new framework of deep CNN is established via designing the convolutional layers and sub-sampling layers. The learning process and results are visualized by t-distributed stochastic neighbor embedding (t-SNE). The results of the experiment present a higher classification accuracy compared with other models. It is demonstrated that the proposed approach is effective and stable for fault diagnosis of a hydraulic axial piston pump.

A Fast and Robust Open-Switch Fault Diagnosis Method for Variable-Speed PMSM System

Traditional open-switch fault diagnosis methods suffer from poor rapidity or robustness. To solve this issue, a new differential current observer-based fault diagnosis method is proposed in this article. With the designed differential observer, fault symptoms (residuals) can be generated and adopted for fault diagnosis easily. Considering that the residuals are sensitive to motor operating condition in the conventional model-based method due to model error, an adaptive fault detection threshold is designed. As a result, the false detection and missed detection caused by the change of working condition can be avoided, and stronger robustness against speed, load, and parameter variations can be achieved with superior rapidity compared with existing methods. Finally, the rapidity and robustness of the proposed fault diagnosis method are verified sufficiently through experimental results.

A Fault Diagnosis Method for Train Plug Doors via Sound Signals

The train plug door is the only way for passengers to get on and off. The reliability of the doors has a direct impact on passengers' safety and operational efficiency. In order to address the shortcomings of the post-analysis and poor real-time of current fault diagnosis methods for train plug doors, a fault diagnosis method based on sound recognition is proposed. To process the non-stationary sound signals, the empirical mode decomposition (EMD) method is applied to sound signal samples of train plug doors, and a series of intrinsic mode functions (IMFs) are obtained. Then, wavelet packet decomposition is utilized on each IMF to acquire more detailed information. And wavelet packet energy entropy features are obtained. The Fisher discrimination criterion is used to carry out a mathematical analysis to select the most significant features as discrimination features. Finally, multi-class support vector machine (multi-class SVM) is utilized to carry out classification and validation. And the prediction accuracy of the 67 test samples reaches 95.52%, which indicates the proposed fault diagnosis method for train plug doors is feasible. The proposed method also provides the possibility of automatic faults recognition.

Current covariance analysis-based open-circuit fault diagnosis for voltage-source-inverter-fed vector-controlled induction motor drives

In recent years, a faulty leg is mostly replaced by a healthy leg for current fault-tolerant control of power converters, and its fault diagnosis only needs to confirm the fault leg. For fault-tolerant control, this paper proposed a real-time, robust and accurate diagnosis method for the legs of insulated gate bipolar transistor open-circuit faults in the pulse-width-modulated voltage source inverters for vector-controlled induction motors based on three-phase current. The covariance of three-phase current is used to diagnose and locate faults by analyzing the correlation between them, and its performance is explained specifically under both health condition and fault conditions. The proposed method is simple and no additional current or voltage sensors are required. The fault diagnosis time is about half a current cycle. Experiments are conducted and the obtained results demonstrate the effectiveness and merit of the proposed method.

Deep balanced domain adaptation neural networks for fault diagnosis of planetary gearboxes with limited labeled data

Due to the time-varying working conditions of planetary gearboxes, the labeled data available are usually limited, current intelligent fault diagnosis methods cannot achieve satisfactory results in the case of limited labeled data. In this paper, a novel method called deep balanced domain adaptation neural network (DBDANN) is proposed for fault diagnosis of planetary gearboxes. First, the multiple convolutional layers were used to extract transferable features layer-by-layer from the raw vibration data of source and the target domain. Then, multi-layer balanced domain adaptation is applied to help training the model, which can reduce the discrepancy of the marginal probability distribution and the conditional probability distribution simultaneously. At last, the test data was fed into the model to evaluate the performance. Experiments with multiple cross-domain tasks under varying speeds and loads demonstrate the effectiveness of the DBDANN, and the proposed model obtained a better performance when compared with the state-of-the-art methods.

Research on the Fault Characteristic of Wind Turbine Generator System Considering the Spatiotemporal Distribution of the Actual Wind Speed

A reliable fault monitoring system is one of the conditions that must be considered in the design of large wind farms today. The most important factor for the fault monitoring should be the accurate diagnosis criteria with sensitive fault characteristics. Most of the current fault diagnosis criteria are obtained based on the average wind speed at the center of the hub which is not in accord with the actual wind condition in nature. So, this paper utilizes an equivalent wind speed (EWS), which can describe the actual wind speed spatiotemporal distribution on the rotor disk area considering the effects of wind shear and tower shadow, to analyze the common mechanical and electrical faults again. Firstly, the EWS model applicable to the 3-blade wind turbines is introduced; then the new fault characteristics of the wind turbine rotor aerodynamic imbalance and the stator winding asymmetry are theoretically analyzed based on the EWS model; finally, the simulation platform is built in Matlab/Simulink for comparison and the simulation result is well consistent with the theory analysis. The aim of this research is to find more accurate fault characteristics and help promoting the healthy development of wind power industry.

Fault diagnosis of air-conditioning refrigeration system based on sparse autoencoder

Abstract To overcome the drawbacks of using supervised learning to extract fault features for classification and low nonlinearity of the features in most of current fault diagnosis of air-conditioning refrigeration system, sparse autoencoder (SAE) is presented to extract fault features that are used as the input to the classifier and to achieve fault diagnosis for air-conditioning refrigeration system. The SAE structure is tuned by adjusting the number of hidden layers and nodes to build the optimal model, which is compared with the fault diagnosis model based on support vector machine. Results indicate that the indexes of the model combined with SAE, such as accuracy, precision and recall, are all improved, especially for the faults with high complexity. Besides, SAE shows high generalization ability with small-scale sample data and high efficiency with large-scale data. Obviously, the use of SAE can effectively optimize the diagnosis performance of the classifier.

Fault Diagnosis Method of Exhaust System of Port Vehicle

For the problem that the current fault diagnosis efficiency of vehicle exhaust system based on cold test is low, this paper proposes a fault diagnosis method of port vehicle exhaust system using principal component analysis method. This method used the historical data aggregating normal working conditions of exhaust system, and found the principal component model of causality among variables in the process of expressing the normal working condition according to the statistical mode. If the real-time monitoring data did not correspond with the principal component model, we could determine the existence of fault in the exhaust system, then used change of each variable of measured data to analyze the variance contribution rate of principal component model, realizing fault diagnosis. Experimental result shows that the fault diagnosis noise of the proposed method is about 50dB lower than that of other methods, and the diagnostic stability and diagnosis rate are improved.Ji, W. and Li, B., 2018. Fault diagnosis method of exhaust system of port vehicle. In: Liu, Z.L. and Mi, C. (eds.), Advances in Sustainable Port and Ocean Engineering. Journal of Coastal Research, Special Issue No. 83, pp. 469–473. Coconut Creek (Florida), ISSN 0749-0208.

Research on Current Monitoring and Fault Diagnosis Technology for Control Rod Drive Mechanism

The control rod drive mechanism (CRDM) is the actuator of the reactor control and nuclear safety protection system. An important measure to prevent the occurrence of control rod stuck, sliding rod and driving failure in the reactor is to effectively monitor the operation status of the CRDM. Based on the analysis of the action principle of the control rod driving mechanism, this paper realizes the current monitoring and fault diagnosis of the control rod driving mechanism by studying the detection method, signal analysis and processing method and fault identification method of the driving mechanism, which lays a technical foundation for the application of the current monitoring and fault diagnosis system of the driving mechanism.

Research on Fault Diagnosis Optimization of Intelligent Acquisition Terminal

Intelligent acquisition terminal is an important medium for power users to collect electric energy data and load data. It realizes centralized collection of electricity information through local and remote communication technology. With the continuous development of communication technology, the acquisition terminal is becoming more and more intelligent and its functions are becoming more and more complex. The content of field equipment fault diagnosis analysis technology is getting higher and higher. This paper combines with the current intelligent acquisition terminal fault diagnosis mode to optimize the research, from the technical and management aspects, makes an analysis of the causes of acquisition anomalies and the lack of means, and puts forward reasonable efficiency improvement suggestions.

Analytical investigation on geared rotor system with multi-body fault based on finite element method

With the rapid development of modern industry, the situation of high speed and heavy load is becoming even more relevant in the gear transmission system. Under the condition of high speed and heavy load, the fault type and frequency of the gear transmission system are gradually increasing. Effective and accurate detection of fault location and fault type is one of the difficult problems in today’s fault diagnosis. In the case of high speed and heavy load, the probability of multi-body fault is greatly increased. However, most of the current fault diagnosis methods are limited to the study of single fault characteristics, and do not take into account the multi body fault. In this paper, the single crack, gear coupling crack, single shaft crack, and gear and shaft coupling crack signal are analyzed by means of the short-time Fourier transform, and the corresponding fault characteristics of different fault types are found out. The modal analysis of the fault state of the gear transmission system is carried out, the structural characteristics of the gear drive are verified, and the influence of the different fault forms on the vibration characteristics of the gear is compared.

An Ensemble Deep Convolutional Neural Network Model with Improved D-S Evidence Fusion for Bearing Fault Diagnosis.

Intelligent machine health monitoring and fault diagnosis are becoming increasingly important for modern manufacturing industries. Current fault diagnosis approaches mostly depend on expert-designed features for building prediction models. In this paper, we proposed IDSCNN, a novel bearing fault diagnosis algorithm based on ensemble deep convolutional neural networks and an improved Dempster–Shafer theory based evidence fusion. The convolutional neural networks take the root mean square (RMS) maps from the FFT (Fast Fourier Transformation) features of the vibration signals from two sensors as inputs. The improved D-S evidence theory is implemented via distance matrix from evidences and modified Gini Index. Extensive evaluations of the IDSCNN on the Case Western Reserve Dataset showed that our IDSCNN algorithm can achieve better fault diagnosis performance than existing machine learning methods by fusing complementary or conflicting evidences from different models and sensors and adapting to different load conditions.

A Robust Observer-Based Method for IGBTs and Current Sensors Fault Diagnosis in Voltage-Source Inverters of PMSM Drives

Permanent magnet synchronous motors (PMSMs) drives using three-phase voltage-source inverters (VSIs) are currently used in many industrial applications. The reliability of VSIs is one of the most important factors to improve the reliability and availability levels of the drive. Accordingly, this paper presents a robust fault diagnostic method for multiple insulated gate bipolar transistors (IGBTs) open-circuit faults and current sensor faults in three-phase PMSM drives. The proposed observer-based algorithm relies on an adaptive threshold for fault diagnosis. Current sensor and open-circuit faults can be distinguished and the faulty sensors and/or power semiconductors are effectively isolated. The proposed technique is robust to machine parameters and load variations. Several simulation and experimental results using a vector-controlled PMSM drive are presented, showing the diagnostic algorithm robustness against false alarms and its effectiveness in both IGBTs and current sensors fault diagnosis.

A new switched current circuit fault diagnosis approach based on pseudorandom test and preprocess by using entropy and Haar wavelet transform

This paper presents a new switched current (SI) circuit fault diagnosis approach based on pseudorandom test and preprocess by using entropy and Haar wavelet transform. The proposed method has the capability to detect and identify faulty transistors in SI circuit by analyzing its time response. The use of pseudorandom sequences as a stimulate signal to SI circuit reduces the cost of testing and the overhead of the test generation circuit, and using entropy and Haar wavelet transform to preprocess the time response for feature extraction drastically improves the fault diagnosis efficiency. For both actual experiment and analysis of switched current filters in Z transform (ASIZ) simulation, a low-pass, a band-pass SI filter and a clock feed-through cancellation circuit have been used as test examples to verify the effectiveness of the proposed method. The result shows that the accuracy of fault recognition achieved is about 100% by analyzing low-frequency approximations entropy and high-frequency details entropy. Therefore, it indicates that the presented method is superior than other methods.

Stator current fault diagnosis of induction motor bearings based on the fast Fourier transform

Monitoring of the technical state of induction motor bearings is one of the important indicators of reliable and efficient operation of industrial electrical drives. This article considers the problems of fault diagnosis of induction motor bearings with a squirrel-cage rotor by stator consumption current. Test results on an experimental test bench with an electromagnetic powder brake are presented and analyzed. Experiments were carried out during tests of the induction motor at the rated mode with an operable bearing and alternately with two faulty bearings with defects at the inner and outer parts of cages. The characteristic harmonic components of stator currents appearing during operation of the induction motor with faulty bearings are presented. It is revealed that an anticipatory diagnosis of induction motors by stator current makes it possible to reveal the air-gap eccentricity due to the operation of an electric motor with a damaged bearing. Control of the technical state of bearings is important both during at the stage of manufacturing machines, in particularly for powerful and crucial mechanisms, and during their use. The basic features of monitoring the bearings are revealed, and recommendations for diagnosis by stator currents are presented.

Fault diagnosis network design for vehicle on-board equipments of high-speed railway: A deep learning approach

With the rapid development of high-speed railways (HSRs) throughout the world, the fault diagnosis systems of vehicle on-board equipments (VOBEs) for high speed trains have received increasing attention. Since the faults of VOBEs in HSRs are usually uncertain and complex, the current fault diagnosis methods are mainly based on manual judgement in real-world operations, which is generally inefficient and insecurity with the big rail traffic data. In this paper, we propose an automated diagnosis network of VOBE for high-speed train via a deep learning approach. First, we propose a mathematical model to formulate the fault diagnosis problem in HSRs, involving the definition of fault evidence vectors and reason vectors by analyzing the real-world fault data that are collected in Wuhan-Guangzhou high speed railway. Then, a deep belief network (DBN) and its training procedures are developed on the basis of Restricted Boltzmann Machine (RBM). Finally, the proposed diagnosis network is trained and validated with real-world data. Furthermore, we compare the DBN-based fault diagnosis network with k-nearest neighbor (KNN) and ANN-BP (artificial neural network with back propagations). The results indicate that, the developed DBN outperforms both KNN and ANN-BP, and improves the accuracy of fault diagnosis for VOBEs to 90–95% in HSRs.

Current Residual Vector-Based Open-Switch Fault Diagnosis of Inverters in PMSM Drive Systems

This paper presents an open-circuit (OC) fault diagnostic method of inverters in closed-loop controlled permanent magnet synchronous motor drive systems based on current residual vector (CRV). This method can get rid of effects of the load and the main controller to obtain reliable detection. The motor drive system is a hybrid system composed of discrete switching signal events and continuous current state variables, which evolves in certain law driven by discrete events. At the base of analyzing operation modes of the system, a mixed logical dynamic (MLD) model of the motor drive system is built to estimate motor currents. Consequently, the proposed fault diagnostic scheme is constructed by the healthy current estimator based on the MLD model in parallel with the plant. When an OC fault occurs, the CRV between the estimator and the plant will show featured amplitude and phase, and hence the fault can be detected. The method utilizes variables already used by the main controller, avoiding the use of extra sensors. During the implementation, amplitude threshold and phase sector are used in order to avoid effects of measurement errors, model parameter errors and noises, which improves the reliability and robustness. Simulation and experimental results validate the proposed diagnostic method, and the diagnostic duration can be reduced within a quarter of fundamental wave period.

Fault Diagnosis Method Based on KPCA and Selective Neural Network Ensemble

Aiming at the problems of less study sample, large network scale and long training time existing in current fault diagnosis field, we develop a new method based on KPCA and selective neural network ensemble. First, reducing the data size by using KPCA to extract the sample features. Then achieving a selective neural network ensemble method based on improved binary particle swarm optimization algorithm (IBPSOSEN), and combining the two methods for fault diagnosis. In selective neural network algorithm, bagging method is used to take a number of different training sets of fault samples to solve the problem of less fault samples. Finally, simulation experiments and comparisons over Tennessee Eastman Process (TE) demonstrate the effectiveness and feasibility of the proposed method.

Fault Diagnosis Mechanism Based on FTA and Bayesian for Large-Scale CNC Equipments

In view of the disadvantage of current FAT-based fault diagnosis method in large-scale complicated system, fault diagnosis method of heavy NC machine based on FTA and Bayesian is discussed. Firstly, building fault trees with the help of reachability matrix, and to set the determinate conditions at every node of fault tree combining FTA with rule reasoning, the minimum cut set of fault reasons are determined as a result of step by step screening fault tree from top-down; Secondly, Bayesian method is integrated into the fault tree diagnostic method to calculate the posterior probability triggered by each fault tree in order to locate the fault tree where the fault had occurred and ensure high efficiency of fault diagnosis; Finally, B/S based intelligent fault diagnosis system for large-scale CNC equipments is developed, and the feasibility and efficiency of this method are proved in an example of fault diagnosis of Φ 160 NC boring and milling machine.

The Fault Prediction of Aerospace Equipment PHM Technology and its Demonstrated Failure Prediction Module Simulation

Prognostics and Health Management Technology (PHM) will make up the deficiencies of current device monitoring and fault diagnosis, especially the lack of fault prediction. Its combination with the neural network can provide a universal theory and technology of the intelligence prediction. Though the experiments, we establish the BP neural network, as well as the most suitable prediction model. After testing data verification, neural networks can accurately predict the status of the equipment, and the health trends in the future. With the network we can accurately predict the system state, remaining life for the aerospace equipment, make it possible to provide maintenance in time, reduce failure losses and improve reliability.