Remaining Useful Life Prediction of Rail Based on Improved Pulse Separable Convolution Enhanced Transformer Encoder

In order to prevent possible casualties and economic loss, it is critical to accurate prediction of the Remaining Useful Life (RUL) in rail prognostics health management. However, the traditional neural networks is difficult to capture the long-term dependency relationship of the time series in the modeling of the long time series of rail damage, due to the coupling relationship of multi-channel data from multiple sensors. Here, in this paper, a novel RUL prediction model with an enhanced pulse separable convolution is used to solve this issue. Firstly, a coding module based on the improved pulse separable convolutional network is established to effectively model the relationship between the data. To enhance the network, an alternate gradient back propagation method is implemented. And an efficient channel attention (ECA) mechanism is developed for better emphasizing the useful pulse characteristics. Secondly, an optimized Transformer encoder was designed to serve as the backbone of the model. It has the ability to efficiently understand relationship between the data itself and each other at each time step of long time series with a full life cycle. More importantly, the Transformer encoder is improved by integrating pulse maximum pooling to retain more pulse timing characteristics. Finally, based on the characteristics of the front layer, the final predicted RUL value was provided and served as the end-to-end solution. The empirical findings validate the efficacy of the suggested approach in forecasting the rail RUL, surpassing various existing data-driven prognostication techniques. Meanwhile, the proposed method also shows good generalization performance on PHM2012 bearing data set.

Characterization,identification and life prediction of acoustic emission signals of tensile damage for HSR gearbox housing material

Purpose-This study aims to ensure the operation safety of high speed trains,it is necessary to carry out nondestructive monitoring of the tensile damage of the gearbox housing material in rail time,yet the traditional tests of mechanical property can hardly meet this requirement.Design/methodology/approach-In this study the acoustic emission(AE)technology is applied in the tensile tests of the gearbox housing material of an high-speed rail(HSR)train,during which the acoustic signatures are acquired for parameter analysis.Afterward,the support vector machine(SVM)classifier is introduced to identify and classify the characteristic parameters extracted,on which basis the SVM is improved and the weighted support vector machine(WSVM)method is applied to effectively reduce the misidentification of the SVM classifier.Through the study of the law of relations between the characteristic values and the tensile life,a degradation model of the gearbox housing material amid tensile is built.Findings-The results show that the growth rate of the logarithmic hit count of AE signals and that of logarithmic amplitude can well characterize the stage of the material tensile process,and the WSVM method can improve the classification accuracy of the imbalanced data to above 94%.The degradation model built can identify the damage occurred to the HSR gearbox housing material amid the tensile process and predict the service life remains.Originality/value-The results of this study provide new concepts for the life prediction of tensile samples,and more further tests should be conducted to verify the conclusion of this research.

Remaining Useful Life Prediction With Partial Sensor Malfunctions Using Deep Adversarial Networks

In recent years,intelligent data-driven prognostic methods have been successfully developed,and good machinery health assessment performance has been achieved through explorations of data from multiple sensors.However,existing datafusion prognostic approaches generally rely on the data availability of all sensors,and are vulnerable to potential sensor malfunctions,which are likely to occur in real industries especially for machines in harsh operating environments.In this paper,a deep learning-based remaining useful life(RUL)prediction method is proposed to address the sensor malfunction problem.A global feature extraction scheme is adopted to fully exploit information of different sensors.Adversarial learning is further introduced to extract generalized sensor-invariant features.Through explorations of both global and shared features,promising and robust RUL prediction performance can be achieved by the proposed method in the testing scenarios with sensor malfunctions.The experimental results suggest the proposed approach is well suited for real industrial applications.

Remaining useful life prediction based on nonlinear random coefficient regression model with fusing failure time data

Remaining useful life(RUL) prediction is one of the most crucial elements in prognostics and health management(PHM). Aiming at the imperfect prior information, this paper proposes an RUL prediction method based on a nonlinear random coefficient regression(RCR) model with fusing failure time data.Firstly, some interesting natures of parameters estimation based on the nonlinear RCR model are given. Based on these natures,the failure time data can be fused as the prior information reasonably. Specifically, the fixed parameters are calculated by the field degradation data of the evaluated equipment and the prior information of random coefficient is estimated with fusing the failure time data of congeneric equipment. Then, the prior information of the random coefficient is updated online under the Bayesian framework, the probability density function(PDF) of the RUL with considering the limitation of the failure threshold is performed. Finally, two case studies are used for experimental verification. Compared with the traditional Bayesian method, the proposed method can effectively reduce the influence of imperfect prior information and improve the accuracy of RUL prediction.

LIFE PREDICTION APPROACH FOR RANDOM MULTIAXIAL FATIGUE

According to the concept of critical plane, a life prediction approach forrandom multiaxial fatigue is presented. First, the critical plane under the multiaxial randomloading is determined based on the concept of the weight-averaged maximum shear strain direction.Then the shear and normal strain histories on the determined critical plane are calculated and takenas the subject of multiaxial load simplifying and multiaxial cycle counting. Furthermore, amultiaxial fatigue life prediction model including the parameters resulted from multiaxial cyclecounting is presented and applied to calculating the fatigue damage generated from each cycle.Finally, the cumulative damage is added up using Miner's linear rule, and the fatigue predictionlife is given. The experiments under multiaxial loading blocks are used for the verification of theproposed method. The prediction has a good correction with the experimental results.

Casing life prediction using Borda and support vector machine methods

Eight casing failure modes and 32 risk factors in oil and gas wells are given in this paper. According to the quantitative analysis of the influence degree and occurrence probability of risk factors, the Borda counts for failure modes are obtained with the Borda method. The risk indexes of failure modes are derived from the Borda matrix. Based on the support vector machine （SVM）, a casing life prediction model is established. In the prediction model, eight risk indexes are defined as input vectors and casing life is defined as the output vector. The ideal model parameters are determined with the training set from 19 wells with casing failure. The casing life prediction software is developed with the SVM model as a predictor. The residual life of 60 wells with casing failure is predicted with the software, and then compared with the actual casing life. The comparison results show that the casing life prediction software with the SVM model has high accuracy.

An Integrated Approach to Fatigue Life Prediction of Whole System for Offshore Platforms

The failure of one or even more components usually does riot lead to the collapse of the whole structure. Most of the analysis of fatigue is centered on only a single component which the researchers are interested in or Much attention should be paid to. However, the collapse of a structure is the result of failure of a series of components in a specific order or path. This paper proposes an integrated approach to fatigue life prediction of whole structural system for offshore platforms, mainly describing the basic principles and prediction method. A method is presented for determining the failure path of the whole structure system and calculating the fatigue life in the determined failure path, The corresponding final collapse criteria for the whole structure system are discussed, A simple method of equivalent fatigue stress range calculation and a mathematical model of structural component fatigue life estimation in consideration of sea wave and sea ice loads are provided. As an application of the proposed approach, a fixed production platform Bohai No. 8 is chosen for the predication of fatigue life of the whole structure system by means of the software OSFAC developed based on the present methods.

Life Prediction Based on Transient Dynamics Analysis of Van Semi-trailer with Air Suspension System

The early fatigue damage in the van-body of the semi-trailer is often caused by the unique mechanical characteristics and the dynamic impact of the loads.The traditional finite element method with static strength analysis cannot support the fatigue design of van-body;thus,the dynamics analysis should be adopted for the endurance performance.The accurate dynamics model to describe the transient impacts of all kinds of uneven road and the proper system transfer functions to calculate the load transfer effects from tire to van-body are two critical factors for transient dynamics analysis.In order to evaluate the dynamic performance,the dynamics model of the trailer with the air suspension is brought forward.Then the analysis method of the power spectral density （PSD） is set up to study the transient responses of the road dynamic impacts.The transient responses transferred from axles to van-body are calculated,such as dynamic stress,dynamic RMS acceleration,and dynamic load factors.Based on the above dynamic responses,the fatigue life of van-body is predicted with the finite element analysis （FEA） method.Applying the test parameters of the trailer with air suspension,the simulation system with Matlab/Simulink is constructed to describe the dynamic responses of the impacts of the tested PSD of the vehicle axles,and then the fatigue life is predicted with FEA method.The simulated results show that the vibration level of the van-body with air suspension is reduced and the fatigue life is improved.The real vehicle tests on different roads are carried out,and the test results validate the accuracy of the simulation system.The proposed fatigue life prediction method is effective for the virtual design of auto-body.

Fatigue Life Prediction of Rolling Bearings Based on Modified SWT Mean Stress Correction

The existing engineering empirical life analysis models are not capable of considering the constitutive behavior of materials under contact loads;as a consequence,these methods may not be accurate to predict fatigue lives of roll-ing bearings.In addition,the contact stress of bearing in operation is cyclically pulsating,it also means that the bear-ing undergo non-symmetrical fatigue loadings.Since the mean stress has great effects on fatigue life,in this work,a novel fatigue life prediction model based on the modified SWT mean stress correction is proposed as a basis of which to estimate the fatigue life of rolling bearings,in which,takes sensitivity of materials and mean stress into account.A compensation factor is introduced to overcome the inaccurate predictions resulted from the Smith,Watson,and Topper(SWT)model that considers the mean stress effect and sensitivity while assuming the sensitivity coefficient of all materials to be 0.5.Moreover,the validation of the model is finalized by several practical experimental data and the comparison to the conventional SWT model.The results show the better performance of the proposed model,especially in the accuracy than the existing SWT model.This research will shed light on a new direction for predicting the fatigue life of rolling bearings.

FATIGUE LIFE PREDICTION THEORY OF COMPOSITE LAMINATES AND EXPERIMENTAL VERIFICATION

According to traditional phenomenological fatigue methodology and moderncontinuum damage mechanics theory, dual fatigue cumulative damage rules to predict fatigue damageformation and propagation lives of the notched composite laminates are presented. A 3-dimensionaldamage constitutive equation of anisotropic composites is also established. Damage strain energyrelease rate is interpreted as a driving force of the fatigue delamination damage propagation. A newdamage evolution equation and a damage propagation σ_a-σ_m-N~* surface (stress amplitude-meanstress-life surface) are derived. Hence, using the method above, the fatigue life of compositecomponents can be predicted. Finally, theoretically predicted results are compared with experimentaldata. It is found that the deviation of theoretic prediction from experimental results is about22%.

Position Encoding Based Convolutional Neural Networks for Machine Remaining Useful Life Prediction

Accurate remaining useful life(RUL)prediction is important in industrial systems.It prevents machines from working under failure conditions,and ensures that the industrial system works reliably and efficiently.Recently,many deep learning based methods have been proposed to predict RUL.Among these methods,recurrent neural network(RNN)based approaches show a strong capability of capturing sequential information.This allows RNN based methods to perform better than convolutional neural network(CNN)based approaches on the RUL prediction task.In this paper,we question this common paradigm and argue that existing CNN based approaches are not designed according to the classic principles of CNN,which reduces their performances.Additionally,the capacity of capturing sequential information is highly affected by the receptive field of CNN,which is neglected by existing CNN based methods.To solve these problems,we propose a series of new CNNs,which show competitive results to RNN based methods.Compared with RNN,CNN processes the input signals in parallel so that the temporal sequence is not easily determined.To alleviate this issue,a position encoding scheme is developed to enhance the sequential information encoded by a CNN.Hence,our proposed position encoding based CNN called PE-Net is further improved and even performs better than RNN based methods.Extensive experiments are conducted on the C-MAPSS dataset,where our PE-Net shows state-of-the-art performance.

Storage life prediction under pre-strained thermally-accelerated aging of HTPB coating using the change of crosslinking density

In order to predict the storage life of a certain type of HTPB(hydroyl-terminated polybutadiene)coating at 25℃ and analyze the influence of pre-strain on the storage life,the accelerated aging tests of HTPB coating at 40℃,50℃,60℃,70℃ with the pre-strain of 0%,3%,6%,9%,respectively were carried out.The variation regularity of the change of crosslinking density was analyzed and the aging model of HTPB coating under pre-strained thermally-accelerated aging was proposed.The storage life of HTPB coating at 25℃ was estimated by using the Berthelot equation as the end point of the aging life with a 30% decrease in maximum elongation.The results showed that the change of crosslinking density of HTPB coating increased with the increase of aging temperature and aging time,and decreased with the increase of pre-strain.Under 0% prestrain,the relationship between the change of crosslinking density of HTPB coating and the aging time can be described by the logarithmic model with the confidence probability greater than 99%.The stress relaxation phenomenon existed under 3%,6%and 9%pre-strained aging.The aging model considering chemical aging and pre-strain was established with the confidence probability greater than 90%.The storage life of HTPB coating was 15.2935 years at 25C under 0% prestrain,which was reduced by 13.9007%,75.6949% and 89.7859% under 3%,6% and 9% pre-strain,respectively.The existence of pre-strain has a serious impact on the storage life of HTPB coating,therefore,the pre-strain should be avoided as much as possible during the actual storage.

Practical Options for Adopting Recurrent Neural Network and Its Variants on Remaining Useful Life Prediction

The remaining useful life(RUL)of a system is generally predicted by utilising the data collected from the sensors that continuously monitor different indicators.Recently,different deep learning(DL)techniques have been used for RUL prediction and achieved great success.Because the data is often time-sequential,recurrent neural network(RNN)has attracted significant interests due to its efficiency in dealing with such data.This paper systematically reviews RNN and its variants for RUL prediction,with a specific focus on understanding how different components(e.g.,types of optimisers and activation functions)or parameters(e.g.,sequence length,neuron quantities)affect their performance.After that,a case study using the well-studied NASA’s C-MAPSS dataset is presented to quantitatively evaluate the influence of various state-of-the-art RNN structures on the RUL prediction performance.The result suggests that the variant methods usually perform better than the original RNN,and among which,Bi-directional Long Short-Term Memory generally has the best performance in terms of stability,precision and accuracy.Certain model structures may fail to produce valid RUL prediction result due to the gradient vanishing or gradient exploring problem if the parameters are not chosen appropriately.It is concluded that parameter tuning is a crucial step to achieve optimal prediction performance.

Bi-variable damage model for fatigue life prediction of metal components

Based on the theory of continuum damage mechanics,a bi-variable damage mechanics model is developed,which,according to thermodynamics,is accessible to derivation of damage driving force,damage evolution equation and damage evolution criteria. Furthermore,damage evolution equations of time rate are established by the generalized Drucker＇s postulate. The damage evolution equation of cycle rate is obtained by integrating the time damage evolution equations,and the fatigue life prediction method for smooth specimens under repeated loading with constant strain amplitude is constructed. Likewise,for notched specimens under the repeated loading with constant strain amplitude,the fatigue life prediction method is obtained on the ground of the theory of conservative integral in damage mechanics. Thus,the material parameters in the damage evolution equation can be obtained by reference to the fatigue test results of standard specimens with stress concentration factor equal to 1,2 and 3.

Life Prediction Model of Machine Tool based on Deep Learning

In view of the shortage of traditional life prediction methods for machine tools,such as low accuracy of life prediction and few samples basis attributes,a life prediction model of machine tools combined with machine tool attributes is proposed.The life prediction model of machine tool adopts KL dispersion distribution theory,uses modal superposition method to carry out machine tool life analysis,calculates the theoretical life of machine tool,and then carries on the simulation,obtains the machine tool life prediction value.Compared with the traditional method of machine tool life prediction,the model is based on the application life fatigue damage model,which superimposes the service times and maintenance cycle of the machine tool,derives the influence factor of machine tool life,and obtains the linear relationship between the influence factor of machine tool life and the life of machine tool.The influence factor of machine tool life is introduced as the life prediction parameter of machine tool.The data transformation relationship of HT300 parts is constructed.The original part data is enhanced.The effective training set is obtained.The life prediction model of machine tool based on deep learning is completed.The quantitative analysis of machine tool life is carried out.The experiment of machine tool life prediction using training data set proves the validity of the model.Regression test was carried out on the training data set to reflect the robustness of the model.The prediction accuracy of the model is further verified by Weibull test.

Energetical Theories on Fatigue Life Prediction

Current developments on fatigue life prediction methods have been systematically reviewed.In con- sideration of the irreversibility of energy dissipation during fatigue damage process,the main contents for fatigue damage estimation and localized equivalence as well as simulation models have been established.A frame of energy-based fatigue life prediction method has been proposed,meanwhile, the procedure in application to a practical structure component has been described.

Dense-Structured Network Based Bearing Remaining Useful Life Prediction System

This work is focused on developing an effective method for bearing remaining useful life predictions.The method is useful in accurately predicting the remaining useful life of bearings so that machine damage,production outage,and human accidents caused by unexpected bearing failure can be prevented.This study uses the bearing dataset provided by FEMTO-ST Institute,Besancon,France.This study starts with the exploration of neural networks,based on which the biaxial vibration signals are modeled and analyzed.This paper introduces pre-processing of bearing vibration signals,neural network model training and adjustment of training data.The model is trained by optimizing model parameters and verifying its performance through cross-validation.The proposed model’s superiority is also confirmed through a comparison with other traditionalmodels.In this study,the neural network model is trained with various types of bearing data and can successfully predict the remaining useful life.The algorithm proposed in this study achieves a prediction accuracy of coefficient of determination as high as 0.99.

Deep Neural Networks Based Approach for Battery Life Prediction

The Internet of Things(IoT)and related applications have witnessed enormous growth since its inception.The diversity of connecting devices and relevant applications have enabled the use of IoT devices in every domain.Although the applicability of these applications are predominant,battery life remains to be a major challenge for IoT devices,wherein unreliability and shortened life would make an IoT application completely useless.In this work,an optimized deep neural networks based model is used to predict the battery life of the IoT systems.The present study uses the Chicago Park Beach dataset collected from the publicly available data repository for the experimentation of the proposed methodology.The dataset is pre-processed using the attribute mean technique eliminating the missing values and then One-Hot encoding technique is implemented to convert it to numerical format.This processed data is normalized using the Standard Scaler technique.Moth Flame Optimization(MFO)Algorithm is then implemented for selecting the optimal features in the dataset.These optimal features are finally fed into the DNN model and the results generated are evaluated against the stateof-the-art models,which justify the superiority of the proposed MFO-DNN model.

Fatigue Life Prediction of Horizontal Press Frame Based on Statistical Probability and Its Redesign

Horizontal press as an important part of hydro-forming machine is used to output the horizontal force to keep the high internal pressure during tube hydro-forming. However,the horizontal press frame is usually mounted on the press bed and not pre-stressed. Meanwhile it will be subjected to the reaction force caused by liquid pressure. Stresses are concentrated severely on the assemble region due to deformation,and total fatigue life will decrease. In order to predict the total fatigue life of the frame,the simulations are used firstly to determine to stress concentration region,and then strain gauge measurements are carried out under different loads. Next,the methods of statistical probability are conducted to calculate the fatigue life based on long-term load history. Finally a structure with the considerable longer fatigue life is redesigned.

High cycle fatigue life prediction method for tail gearbox casing of a helicopter transmission system

A method and procedure of high cycle fatigue life prediction for helicopter transmission system tail gearbox casing is presented, including fatigue test load, three parameters S-N curve, reduction factor and cumulative damage law. According to the fatigue test results, the design load spectrum and the three parameters S-N curve, a fatigue life prediction of the tail gearbox casing of a helicopter is performed as an example.