摘要:Diagnostics and condition monitoring are aspects of a field of engineering that aims to manage the health of engineering machinery and structures,but which has increasing applicability to many other complex systems in the world from the medical care of people,through the reliability of supply chain logistics to the secure operation of a large and complex organisation.The management of health generally requires a number of sequential steps,which are usually as follows:the detection of any abnormality in the procedural operation of the system,the location of the abnormal behaviour within the system,an assessment of the severity of the abnormality and identification of the potential system vulnerabilities that result from it,a detailed diagnosis of the problem including the identification of its root cause,and finally a predictive assessment of the future prospects for the continued operation of the system,including any remedial action that should be taken to minimise impact and maximise continued operational performance.
摘要:Recently,deep learning(DL)has been widely used in the field of remaining useful life(RUL)prediction.Among various DL technologies,recurrent neural network(RNN)and its variant,e.g.,long short-term memory(LSTM)network,have gained extensive attention for their ability to capture temporal dependence.Although existing RNN-based methods have demonstrated their RUL prediction effectiveness,they still suffer from the following two limitations:1)it is difficult for the RNN to directly extract degradation features from original monitoring data and 2)most RNN-based prognostics methods are unable to quantify RUL uncertainty.To address the aforementioned limitations,this paper proposes a new prognostics method named residual convolution LSTM(RC-LSTM)network.In the RC-LSTM,a new ResNet-based convolution LSTM(Res-ConvLSTM)layer is stacked with a convolution LSTM(ConvLSTM)layer to extract degradation representations from monitoring data.Then,under the assumption that the RUL follows a normal distribution,an appropriate output layer is constructed to quantify the uncertainty of prediction results.Finally,the effectiveness and superiority of the RC-LSTM are verified using monitoring data from accelerated bearing degradation tests.
摘要:Modulation of the gear mesh vibration is a major field of research for the condition monitoring of planetary gearboxes.The modulation creates sidebands around the gearmesh frequency in the vibration spectrum,and the distribution of these sidebands has been researched in numerous papers.All publications on the subject assume that the effect of the time varying signal propagation delay between the main vibration source–the gear mesh point(s)–and the(usually fixed)transducer can be neglected.This paper investigates the validity of this assumption.To do so,a planetary gearbox with a transducer mounted on the(fixed)ring gear is studied,and the effect of the propagation delay is modelled as a phase modulation of the gear mesh vibration.General expressions are then derived for the distribution and strength of the modulation sidebands,and these expressions are applied to quantify the effect of the propagation delay on five industrial gearboxes.The results show that the amplitude of the sidebands is negligible and would not interfere with condition assessment based on analysis of the modulation of the gear mesh frequency,and thus the propagation delay can be neglected for practical purposes.
摘要:Time-frequency methods are effective tools in identifying the frequency content of a signal and revealing its timevariant features.This paper presents the use of instantaneous features(i.e.,instantaneous energy and signal phase)of acoustic emission(AE)in the detection of thermal damage to the workpiece in grinding.The low-order frequency moments of a scalogram are used to obtain both the instantaneous energy and the mean frequency at which the signal phase is recovered.The grinding process is monitored using AE for a variety of operating conditions,including regular grinding,grinding at higher cutting speed and larger feed,and small dressing depth of cut.The instantaneous features extracted by the scalogram are compared with the results obtained by the empirical mode decomposition.It has been found that both the instantaneous energy and phase deviation indicate the presence of burn damage and serve as robust and reliable indicators,providing a basis for detecting the grinding burn.
摘要:Submission template For our Microsoft Word submission template,please download the file form the following link:https://ojs.istp-press.com/dmd/about/submissions.Manuscript presentation 1.Length:An original article would normally consist of 4000-6000 words(excluding figures,tables and references),although high-quality articles which exceed 7000 words will be considered.Review papers which summarize current research activities should have a minimum length limit of 10000 words(excluding figures,tables and references).
摘要:Thanks to the fast development of micro-electro-mechanical systems(MEMS)technologies,MEMS accelerometers show great potentialities for machine condition monitoring.To overcome the problems of a poor signal to noise ratio(SNR),complicated modulation,and high costs of vibration measurement and computation using conventional integrated electronics piezoelectric accelerometers,a triaxialMEMS accelerometer-based on-rotor sensing(ORS)technology was developed in this study.With wireless data transmission capability,the ORS unit can be mounted on a rotating rotor to obtain both rotational and transverse dynamics of the rotor with a high SNR.The orthogonal outputs lead to a construction method of analytic signals in the time domain,which is versatile in fault detection and diagnosis of rotating machines.Two case studies based on an induction motor were carried out,which demonstrated that incipient bearing defect and half-broken rotor bar can be effectively diagnosed by the proposed measurement and analysis methods.Comparatively,vibration signals from translational on-casing accelerometers are less capable of detecting such faults.This demonstrates the superiority of the ORS vibrations in fault detection of rotating machines.
摘要:In this work,a comparative study is performed to investigate the influence of time-varying normal forces on the friction properties and friction-induced stick-slip vibration(FIV)by experimental and theoretical methods.In the experiments,constant and harmonic-varying normal forces are applied,respectively.The measured vibration signals under two loading forms are compared in both time and frequency domains.In addition,mathematical tools such as phase space reconstruction and Fourier spectra are used to reveal the science behind the complicated dynamic behavior.It can be found that the friction system shows steady stick-slip vibration,and the main frequency does not vary with the magnitude of the constant normal force,but the size of limit cycle increases with the magnitude of the constant normal force.In contrast,the friction system under the harmonic normal force shows complicated behavior,for example,higher-frequency larger-amplitude vibration occurs and looks chaotic as the frequency of the normal force increases.The interesting findings offer a new way for controlling FIV in engineering applications.
摘要:In this paper,a novel time–frequency(TF)analysis method,called the short-time Fourier transform using odd symmetric window function(OSTFT),is proposed using odd symmetric window function to replace the conventional even window function of STFT.Different from conventional STFT acquiring the amplitude maximum at time and frequency centers,OSTFT acquires the minimum amplitude of 0.Hence,OSTFT can obtain a TF representation(TFR)with high TF resolution by utilizing the leaked energy rather than restraining it.It is worth to mention that the proposed OSTFT can vitiate the effect of window size we choose on the TFR obtained.Furthermore,it also has a good performance on signals with complex instantaneous frequencies(IFs),even crossing IFs.Because we just replace the conventional window function of STFT,the time-consuming of the proposed OSTFT is at the same level as the conventional STFT.The effectiveness of proposed OSTFT has been validated on two complex multi-component simulated numerical signals and a signal collected from the brown bat.
摘要:Modern manufacturing aims to reduce downtime and track process anomalies to make profitable business decisions.This ideology is strengthened by Industry 4.0,which aims to continuously monitor high-value manufacturing assets.This article builds upon the Industry 4.0 concept to improve the efficiency of manufacturing systems.The major contribution is a framework for continuous monitoring and feedback-based control in the friction stir welding(FSW)process.It consists of a CNC manufacturing machine,sensors,edge,cloud systems,and deep neural networks,all working cohesively in real time.The edge device,located near the FSW machine,consists of a neural network that receives sensory information and predicts weld quality in real time.It addresses time-critical manufacturing decisions.Cloud receives the sensory data if weld quality is poor,and a second neural network predicts the new set of welding parameters that are sent as feedback to the welding machine.Several experiments are conducted for training the neural networks.The framework successfully tracks process quality and improves the welding by controlling it in real time.The system enables faster monitoring and control achieved in less than 1 s.The framework is validated through several experiments.
摘要:Integrated with sensors,processors,and radio frequency(RF)communication modules,intelligent bearing could achieve the autonomous perception and autonomous decision-making,guarantying the safety and reliability during their use.However,because of the resource limitations of the end device,processors in the intelligent bearing are unable to carry the computational load of deep learning models like convolutional neural network(CNN),which involves a great amount of multiplicative operations.To minimize the computation cost of the conventional CNN,based on the idea of AdderNet,a 1-D adder neural network with a wide first-layer kernel(WAddNN)suitable for bearing fault diagnosis is proposed in this paper.The proposed method uses the l1-norm distance between filters and input features as the output response,thus making the whole network almost free of multiplicative operations.The whole model takes the original signal as the input,uses a wide kernel in the first adder layer to extract features and suppress the high frequency noise,and then uses two layers of small kernels for nonlinear mapping.Through experimental comparison with CNN models of the same structure,WAddNN is able to achieve a similar accuracy as CNN models with significantly reduced computational cost.The proposed model provides a new fault diagnosis method for intelligent bearings with limited resources.
摘要:This paper proposes an intelligent process fault diagnosis system through integrating the techniques of Andrews plot and convolutional neural network.The proposed fault diagnosis method extracts features from the on-line process measurements using Andrews function.To address the uncertainty of setting the proper dimension of extracted features in Andrews function,a convolutional neural network is used to further extract diagnostic information from the Andrews function outputs.The outputs of the convolutional neural network are then fed to a single hidden layer neural network to obtain the final fault diagnosis result.The proposed fault diagnosis system is compared with a conventional neural network based fault diagnosis system and integrating Andrews function with neural network and manual selection of features in Andrews function outputs.Applications to a simulated CSTR process show that the proposed fault diagnosis system gives much better performance than the conventional neural network based fault diagnosis system and manual selection of features in Andrews function outputs.It reveals that the use of Andrews function and convolutional neural network can improve the diagnosis performance.
摘要:Multiple-stage steam turbine generators,like those found in nuclear power plants,pose special challenges with regards to mechanical unbalance diagnosis.Several factors contribute to a complex vibrational response,which can lead to incorrect assessments if traditional condition monitoring strategies are used without considering the mechanical system as a whole.This,in turn,can lead to prolonged machinery downtime.Several machine learning techniques can be used to integrally correlate mechanical unbalance along the shaft with transducer signals from rotor bearings.Unfortunately,this type of machinery has scarce data regarding faulty behavior.However,a variety of fault conditions can be simulated in order to generate these data using computational models to simulate the dynamic response of individual machines.In the present work,a multibody model of a 640MWsteam turbine flexible rotor is employed to simulate mechanical unbalance in several positions along the shaft.Synchronous components of the resulting vibration signals at each bearing are obtained and utilized as training data for two regression models designed for mechanical unbalance diagnosis.The first approach uses an artificial neural network and the second one utilizes a support vector regression algorithm.In order to test their performance,the stiffness of each bearing in the multibody simulation was altered between 50%and 150%of the training model values,random noise was added to the signal and several dynamic unbalance conditions were simulated.Results show that both approaches can reliably diagnose dynamic rotor unbalance even when there is a typical degree of uncertainty in bearing stiffness values.
摘要:Aiming at the difficulty of mining fault prognosis starting points and constructing prognostic models for remaining useful life(RUL)prediction of rolling bearings,a RUL prediction method is proposed based on health indicator(HI)extraction and trajectory-enhanced particle filter(TE-PF).By extracting a HI that can accurately track the trending of bearing degradation and combining it with the early fault enhancement technology,early abnormal sample nodes can be mined to provide more samples with fault information for the construction and training of subsequent prediction models.Aiming at the problem that traditional degradation rate models based on PF are vulnerable to HI mutations,a TE-PF prediction method is proposed based on comprehensive utilization of historical degradation information to timely modify prediction model parameters.Results from a rolling bearing prognostic study show that prediction starting points can be accurately detected and a reasonable prediction model can be conveniently constructed by the RUL prediction method based on HI amplitude abnormal detection and TE-PF.Furthermore,aiming at the RUL prediction problem under the condition of HI mutation,RUL prediction with probability and statistics characteristics under a confidence interval can be obtained based on the method proposed.
摘要:Scientific research frequently involves the use of computational tools and methods.Providing thorough documentation,open-source code,and data–the creation of reproducible computational research(RCR)–helps others understand a researcher’s work.In this study,we investigate the state of reproducible computational research,broadly,and from within the field of prognostics and health management(PHM).In a text mining survey of more than 300 articles,we show that fewer than 1%of PHM researchers make their code and data available to others.To promote the RCR further,our work also highlights several personal benefits for those engaged in the practice.Finally,we introduce an open-source software tool,called PyPHM,to assist PHM researchers in accessing and preprocessing common industrial datasets.
摘要:Most accidents of centrifugal compressors are caused by fluid pulsation or unsteady fluid excitation.Rotating stall,as an unstable flow phenomenon in the compressor,is a difficult point in the field of fluid machinery research.In this paper,a stack denoising kernel autoencoder neural network method is proposed to study the early warning of rotating stall in a centrifugal compressor.By collecting the pressure pulsation signals of the centrifugal compressor under different flow rates in engineering practice,a double hidden layer sparse denoising autoencoder neural network is constructed.According to the output labels of the network,it can be judged whether the rotation stall occurs.At the same time,the Gaussian kernel is used to optimize the loss function of the whole neural network to improve the signal feature learning ability of the network.From the experimental results,it can be seen that the flow state of the centrifugal compressor is accurately judged,and the rotation stall early warning of the centrifugal compressor at different speeds is realized,which lays a foundation for the research of intelligent operation and maintenance of the centrifugal compressor.
摘要:Surface defects,including dents,spalls,and cracks,for rolling element bearings are the most common faults in rotating machinery.The accurate model for the time-varying excitation is the basis for the vibration mechanism analysis and fault feature extraction.However,in conventional investigations,this issue is not well and fully addressed from the perspective of theoretical analysis and physical derivation.In this study,an improved analytical model for time-varying displacement excitations(TVDEs)caused by surface defects is theoretically formulated.First and foremost,the physical mechanism for the effect of defect sizes on the physical process of rolling element-defect interaction is revealed.According to the physical interaction mechanism between the rolling element and different types of defects,the relationship between time-varying displacement pulse and defect sizes is further analytically derived.With the obtained time-varying displacement pulse,the dynamic model for the deep groove bearings considering the internal excitation caused by the surface defect is established.The nonlinear vibration responses and fault features induced by surface defects are analyzed using the proposed TVDE model.The results suggest that the presence of surface defects may result in the occurrence of the dual-impulse phenomenon,which can serve as indexes for surface-defect fault diagnosis.
摘要:In the field of data-driven bearing fault diagnosis,convolutional neural network(CNN)has been widely researched and applied due to its superior feature extraction and classification ability.However,the convolutional operation could only process a local neighborhood at a time and thus lack the ability of capturing long-range dependencies.Therefore,building an efficient learning method for long-range dependencies is crucial to comprehend and express signal features considering that the vibration signals obtained in a real industrial environment always have strong instability,periodicity,and temporal correlation.This paper introduces nonlocal mean to the CNN and presents a 1D nonlocal block(1D-NLB)to extract long-range dependencies.The 1D-NLB computes the response at a position as a weighted average value of the features at all positions.Based on it,we propose a nonlocal 1D convolutional neural network(NL-1DCNN)aiming at rolling bearing fault diagnosis.Furthermore,the 1D-NLB could be simply plugged into most existing deep learning architecture to improve their fault diagnosis ability.Under multiple noise conditions,the 1D-NLB improves the performance of the CNN on the wheelset bearing data set of high-speed train and the Case Western Reserve University bearing data set.The experiment results show that the NL-1DCNN exhibits superior results compared with six state-of-the-art fault diagnosis methods.
摘要:I.INTRODUCTION AND SCOPE The rapid development of data science and associated artificial intelligence(AI)methods has seen a substantial increase in interest in their application to anomaly detection,fault diagnostic,and prognostic challenges across a wide range of industrial and civil applications.
摘要:This work presents a novel wavelet-based denoising technique for improving the signal-to-noise ratio(SNR)of nonsteady vibration signals in hardware redundant systems.The proposed method utilizes the relationship between redundant hardware components to effectively separate fault-related components from the vibration signature,thus enhancing fault detection accuracy.The study evaluates the proposed technique on two mechanically identical subsystems that are simultaneously controlled under the same speed and load inputs,with and without the proposed denoising step.The results demonstrate an increase in detection accuracy when incorporating the proposed denoising method into a fault detection system designed for hardware redundant machinery.This work is original in its application of a new method for improving performance when using residual analysis for fault detection in hardware redundant machinery configurations.Moreover,the proposed methodology is applicable to nonstationary equipment that experiences changes in both speed and load.
摘要:Bearings are key components in rotating machinery,which is widely used in many fields,such as CNC machines,wind turbines and induction machines.The increasingly harsh operation environment can lead to wear and tear on raceways and reduce the precision and reliability of bearing or even machinery.Lubrication could relieve the wear to some degree,which is benefit to prolong the bearing’s life.Thus,investigation on the vibration responses under the influence of oil film is of great significance.However,for mechanism analysis,how to include the oil film into the bearing dynamic model affects the result and efficiency of solution.To address this problem,this study proposed a fast algorithm through load distribution and interpolation when calculating oil film stiffness and thickness during the solution of bearing vibration model.Analysis of oil film on vibration is carried out and a bearing test rig is designed to verify the proposed model.Numerical simulation result shows that rotational speed and load have vital effect on oil film and vibration.The experimental result is consistent with the simulation,which shows that the proposed model has a better performance on modeling bearing vibration and the method of considering oil film is reasonable.
摘要:Sensing is the fundamental technique for sensor data acquisition in monitoring the operation condition of the machinery,structures,and manufacturing processes.In this paper,we briefly discuss the general idea and advances of various new sensing technologies,including multiphysics sensing,smart materials and metamaterials sensing,microwave sensing,fiber optic sensors,and terahertz sensing,for measuring vibration,deformation,strain,acoustics,temperature,spectroscopic,etc.Based on the observations from the state of the art,we provide comprehensive discussions on the possible opportunities and challenges of these new sensing technologies so as to steer future development.
摘要:Machine learning,especially deep learning,has been highly successful in data-intensive applications;however,the performance of these models will drop significantly when the amount of the training data amount does not meet the requirement.This leads to the so-called few-shot learning(FSL)problem,which requires the model rapidly generalize to new tasks that containing only a few labeled samples.In this paper,we proposed a new deep model,called deep convolutional meta-learning networks,to address the low performance of generalization under limited data for bearing fault diagnosis.The essential of our approach is to learn a base model from the multiple learning tasks using a support dataset and finetune the learnt parameters using few-shot tasks before it can adapt to the new learning task based on limited training data.The proposed method was compared to several FSL methods,including methods with and without pre-training the embedding mapping,and methods with finetuning the classifier or the whole model by utilizing the few-shot data from the target domain.The comparisons are carried out on 1-shot and 10-shot tasks using the Case Western Reserve University bearing dataset and a cylindrical roller bearing dataset.The experimental result illustrates that our method has good performance on the bearing fault diagnosis across various few-shot conditions.In addition,we found that the pretraining process does not always improve the prediction accuracy.
摘要:Anomaly detection(AD)is an important task in a broad range of domains.A popular choice for AD are Deep Support Vector Data Description models.When learning such models,normal data is mapped close to and anomalous data is mapped far from a center,in some latent space,enabling the construction of a sphere to separate both types of data.Empirically,it was observed:(i)that the center and radius of such sphere largely depend on the training data and model initialization which leads to difficulties when selecting a threshold,and(ii)that the center and radius of this sphere strongly impact the model AD performance on unseen data.In this work,a more robust AD solution is proposed that(i)defines a sphere with a fixed radius and margin in some latent space and(ii)enforces the encoder,which maps the input to a latent space,to encode the normal data in a small sphere and the anomalous data outside a larger sphere,with the same center.Experimental results indicate that the proposed algorithm attains higher performance compared to alternatives,and that the difference in size of the two spheres has a minor impact on the performance.
摘要:In this paper,we present an alternative technique for detecting changes in the operating conditions of rolling element bearings(REBs)that can lead to premature failure.The developed technique is based on measuring the kinematics of the bearing cage.The rotational motion of the cage is driven by traction forces generated in the contacts of the rolling elements with the races.It is known that the cage angular frequency relative to shaft angular frequency depends on the bearing load,the bearing speed,and the lubrication condition since these factors determine the lubricant film thickness and the associated traction forces.Since a large percentage of REB failures are due to misalignment or lubrication problems,any evidence of these conditions should be interpreted as an incipient fault.In this paper,a novel method for the measurement of the instantaneous angular speed(IAS)of the cage is developed.The method is evaluated in a deep groove ball bearing test rig equipped with a cage IAS sensor,as well as a custom acoustic emission(AE)transducer and a piezoelectric accelerometer.The IAS of the cage is analyzed under different bearing loads and shaft speeds,showing the dependence of the cage angular speed with the calculated lubricant film thickness.Typical bearing faulty operating conditions(mixed lubrication regime,lubricant depletion,and misalignment)are recreated.It is shown that the cage IAS is dependent on the lubrication regime and is sensitive to misalignment.The AE signal is also used to evaluate the lubrication regime.Experimental results suggest that the proposed technique can be used as a condition monitoring tool in industrial environments to detect abnormal REB conditions that may lead to premature failure.
摘要:Data-driven methods are widely considered for fault diagnosis in complex systems.However,in practice,the between-class imbalance due to limited faulty samples may deteriorate their classification performance.To address this issue,synthetic minority methods for enhancing data have been proved to be effective in many applications.Generative adversarial networks(GANs),capable of automatic features extraction,can also be adopted for augmenting the faulty samples.However,the monitoring data of a complex system may include not only continuous signals but also discrete/categorical signals.Since the current GAN methods still have some challenges in handling such heterogeneous monitoring data,a Mixed Dual Discriminator GAN(noted as M-D2GAN)is proposed in this work.In order to render the expanded fault samples more aligned with the real situation and improve the accuracy and robustness of the fault diagnosis model,different types of variables are generated in different ways,including floating-point,integer,categorical,and hierarchical.For effectively considering the class imbalance problem,proper modifications are made to the GAN model,where a normal class discriminator is added.A practical case study concerning the braking system of a high-speed train is carried out to verify the effectiveness of the proposed framework.Compared to the classic GAN,the proposed framework achieves better results with respect to F-measure and G-mean metrics.
摘要:In machining processes,chatter vibrations are always regarded as one of the major limitations for production quality and efficiency.Accurate and timely monitoring of chatter is helpful to maintain stable machining operations.At present,most chatter monitoring methods are based on the energy level at specified chatter frequencies or frequency bands.However,the spectral features of chatter could change during machining operations due to complexity and time-varying dynamics of the physical machining process.The purpose of this paper is to investigate the time-varying chatter features in turning of thin-walled tubular workpieces from the perspective of entropy.The airborne acoustics was selected as the source of information for machining condition monitoring.First,corresponding to the distinguishing surface topographies relevant to machining conditions,the features of the sound signal emitted during turning of the thin-walled cylindrical workpieces were extracted using the spectral analysis and wavelet packet transform,respectively.It was shown that the dominant vibration frequency as well as the energy distribution could shift with the transition of the machining status.After that,two relative entropy indicators based on the spectrum and the wavelet packet energy were constructed to identify chattering events in turning of the thin-walled tubes.The experimental results demonstrate that the proposed indicators could accurately reflect the transition of machining conditions with high sensitivity and robustness in comparison with the traditional FFT-based methods.The achievement of this study lays the foundations of the online chatter monitoring and control technique for turning of the thin-walled tubular workpieces.
摘要:Early fault diagnosis of bearings is crucial for ensuring safe and reliable operations.Convolutional neural networks(CNNs)have achieved significant breakthroughs in machinery fault diagnosis.However,complex and varying working conditions can lead to inter-class similarity and intra-class variability in datasets,making it more challenging for CNNs to learn discriminative features.Furthermore,CNNs are often considered“black boxes”and lack sufficient interpretability in the fault diagnosis field.To address these issues,this paper introduces a residual mixed domain attention CNN method,referred to as RMA-CNN.This method comprises multiple residual mixed domain attention modules(RMAMs),each employing one attention mechanism to emphasize meaningful features in both time and channel domains.This significantly enhances the network’s ability to learn fault-related features.Moreover,we conduct an in-depth analysis of the inherent feature learning mechanism of the attention module RMAM to improve the interpretability of CNNs in fault diagnosis applications.Experiments conducted on two datasets—a high-speed aeronautical bearing dataset and a motor bearing dataset—demonstrate that the RMA-CNN achieves remarkable results in diagnostic tasks.
摘要:Multi-sensor measurement iswidely employed in rotatingmachinery to ensure the safety ofmachines.The information provided by the single sensor is not comprehensive.Multi-sensor signals can provide complementary information in characterizing the health condition of machines.This paper proposed a multi-sensor fusion convolution neural network(MF-CNN)model.The proposed model adds a 2-D convolution layer before the classical 1-D CNN to automatically extract complementary features of multi-sensor signals and minimize the loss of information.A series of experiments are carried out on a rolling bearing test rig to verify the model.Vibration and sound signals are fused to achieve higher classification accuracy than typical machine learning model.In addition,the model is further applied to gas turbine abnormal detection,and shows great robustness and generalization.
摘要:The transient impulse features caused by rolling bearing faults are often present in the resonance frequency band which is closely related to the dynamic characteristics of the machine structure.Informative frequency band identification is a crucial prerequisite for envelope analysis and thereby accurate fault diagnosis of rolling bearings.In this paper,based on the ratio of quasi-arithmetic means and Gini index,improved Gini indices(IGIs)are proposed to quantify the transient impulse features of a signal,and their effectiveness and advantages in sparse quantification are confirmed by simulation analysis and comparisons with traditional sparsity measures.Furthermore,an IGI-based envelope analysis method named IGIgram is developed for fault diagnosis of rolling bearings.In the new method,an IGI-based indicator is constructed to evaluate the impulsiveness and cyclostationarity of the narrow-band filtered signal simultaneously,and then a frequency band with abundant fault information is adaptively determined for extracting bearing fault features.The performance of the IGIgram method is verified on the simulation signal and railway bearing experimental signals and compared with typical sparsity measures-based envelope analysis methods and log-cycligram.The results demonstrate that the proposed IGIs are efficient in quantifying bearing fault-induced transient features and the IGIgram method with appropriate power exponent can effectively achieve the diagnostics of different axle-box bearing faults.
摘要:Effective fault diagnosis of planetary gearboxes is critical for ensuring the safety and dependability of mechanical drive systems.Nevertheless,variable conditions and inadequate fault data bring huge challenges to its practical fault diagnosis.Taking this into account,this study presents a new intelligent fault diagnosis(IFD)approach for planetary gearbox using a transferable deep Q network(TDQN)that merges deep reinforcement learning(DRL)and transfer learning(TL).First,a DRL environment simulation is designed by a predefined classification Markov decision process.Then,leveraging varied-size convolutions and residual learning,a multiscale residual convolutional neural network agent for TDQN is created to automatically learn meaningful features directly from vibration signals while avoiding model degradation.Next,a large source dataset is obtained from complex conditions,and this agent learns an IFD policy via autonomous interaction with the data environment.Finally,a parameter-based TL strategy is adopted to retrain the model on target datasets with variable conditions and small training data,which is conducted by fine-tuning the model parameters gained from the source task to accomplish target tasks.The results show that this TDQN outperforms not only state-of-the-art methods in a source task with an accuracy of 98.53%but also in two target tasks with 99.63%and 98.37%,respectively.