Life Prediction Research Articles

A novel sparse Gaussian process regression with time-aware spatiotemporal kernel for remaining useful life prediction and uncertainty quantification of bearings

A novel sparse Gaussian process regression with time-aware spatiotemporal kernel for remaining useful life prediction and uncertainty quantification of bearings

Fatigue life prediction of film-cooling Hole specimens with initial damage

This study investigates a Nickel-based single crystal (SX) superalloy with femtosecond laser-drilled film-cooling holes (FCHs) under varying temperatures (room temperature, 850 °C, and 980 °C), employing a novel framework for predicting fatigue life based on initial manufacturing damage quantification. For all tested anisotropic SX superalloy specimens (including smooth and FCH specimens), the initial damage state is characterized as an equivalent initial flaw size (EIFS), and an EIFS calculation model considering stress concentration is established. Subsequently, the fatigue crack paths and microstructural characteristics of the FCH specimens at different temperatures are analyzed, elucidating crack initiation mechanisms and propagation patterns. A novel incremental plasticity J-integral driving force for fatigue crack propagation is introduced. By incorporating the closure effect of small crack propagation and employing Markov Chain Monte Carlo simulations for determining crack growth rate probabilities, a more accurate expression for the crack growth rate in relation to ΔJfat − ΔJth is derived. This expression comprehensively captures crack patterns on crystallographic planes and Type I mixed mode behavior. Finally, the total fatigue life of the FCH structures, featuring a threefold dispersion zone in both room and high-temperature environments, is predicted through experimental observations and description of crack growth rates. The predicted outcomes significantly outperform those of the conventional life prediction models reliant on crystal plasticity theory.

A residual service life prediction of lithium-ion batteries based on decomposition algorithm and fully connected neural network

A residual service life prediction of lithium-ion batteries based on decomposition algorithm and fully connected neural network

A dual-stream temporal convolutional network for remaining useful life prediction of rolling bearings

Abstract Remaining useful life (RUL) prediction plays an indispensable role in the reliable operation and improved maintenance of rolling bearings. Currently, data-driven methods based on deep learning have made significant progress in RUL prediction. However, most of such methods only consider the correlation between channels, ignoring the importance of different time steps for RUL prediction. In addition, it is still challenging to effectively fuse the degradation features of rolling bearings to improve the model RUL prediction performance. To address the above issues, this paper proposes a novel data-driven RUL prediction method named dual-stream temporal convolution network (DSTCN). First, a hybrid attention temporal convolution block (HATCB) is designed to capture the correlation of degraded features on the channel dimension and temporal dimension. Second, a one-dimensional attention fusion module is designed. This module is capable of weight recalibration and assignment to adaptively fuse different degraded features. Afterward, the Hilbert Marginal spectrum is obtained using the Hilbert Huang Transform and used as the input to one stream. Meanwhile, vibration signals are used as the input of the other stream, thus building a dual-stream temporal convolution network to realize RUL prediction. The effectiveness of the proposed method is validated with two life-cycle datasets, and the results show that the method has lower prediction error than other methods for RUL prediction and prognostic analysis.

Fatigue Life Prediction for Stud Shear Connectors Based on a Machine Learning Model

The shear connector of a steel composite bridge is designed by predicting fatigue life using the fatigue strength curves (S-N curve) based on push-out test results. The fatigue strength curves of the current design codes present only a linear relationship between the stress range and fatigue life on a log scale based on push-out experiment results. However, an alternative to the current empirical formula is necessary for the fatigue design of shear connections involving many detailed variations or high strength steel materials. This study collected and reanalyzed data from push-out fatigue tests to determine the factors influencing fatigue life and propose a machine learning-based fatigue life prediction model. The proposed machine learning model demonstrated an improvement in predictive performance of approximately 2 to 8 times compared to the existing design codes when evaluated against experimental data. Feature importance analysis based on the proposed model revealed that the stress range most significantly influenced fatigue life prediction. Model validation results indicated that the proposed model provided reliable predictions with accuracy and generalization performance. Moreover, it effectively accounted for uncertainty by incorporating features previously overlooked in existing design codes. Plans for fine-tuning pretrained models were also discussed.

Remaining useful life prediction method for jointless track circuits based on multivariate feature fusion and nonlinear Wiener process

Abstract Predicting the Remaining Useful Life (RUL) of track circuits is essential to ensure the safe and reliable operation of high-speed railways. In response to the challenges faced by current machine-learning-based RUL prediction methods, which struggle to represent the uncertainty in the probability distribution of RUL predictions, this paper suggests a hybrid-driven method for estimating remaining life. Firstly, the track circuit Health Index (HI) is constructed by feature dimensionality reduction and fusion of the original multivariate monitoring data through Kernel Principal Component Analysis (KPCA) and Autoencoder (AE); Secondly, the degraded state of the rail circuit is modelled using a nonlinear Wiener degradation model. Finally, the principle of First Hitting Time (FHT) is used to derive the Probability Density Function (PDF) of the anticipated RUL. The efficacy and superiority of the approach presented in this paper are validated by experimental research on the track circuit monitoring dataset. The method enhances forecast accuracy and reduces prediction uncertainty, offering robust technical support for track circuit maintenance decision-making.

Modified crystal plasticity constitutive model considering tensorial properties of microstructural evolution and creep life prediction model for Ni-based single crystal superalloy with film cooling hole

Accurate assessment of the creep life of film cooling hole structures is critical for long-life design and safe operation of aero engines and gas turbines. Firstly, through the high temperature creep experiment of nickel-based single crystal superalloy with film cooling hole, the microstructure evolution process under multiaxial stress state around film hole is characterized. Then, considering the directional effect of rafting structure and the influence of multiaxial stress, a fourth-order tensor is used to describe the evolution of γ phase width, and the microstructure evolution model accounting for multi-axial stress states is established. The microstructure evolution is coupled into the crystal plasticity constitutive model by Orowan stress. Meanwhile, based on continuous damage mechanics, a new multiaxial damage evolution law is established by introducing a multiaxial ductility factor into the constitutive model. The improved crystal plasticity constitutive model can effectively predict the microstructural evolution under multiaxial stress conditions. Furthermore, the combination of the modified crystal plasticity constitutive model and the critical distance method considering stress gradients is used for life prediction of film cooling hole structures. The prediction results show the effectiveness and necessity of considering the microstructure evolution in the life prediction.

Bearing remaining life prediction method based on ARAD -ELN and multi-stage wiener process

Abstract Stochastic process-based models are extensively utilized in health assessments and Remaining Useful Life (RUL) predictions of bearings. Nevertheless, bearings in actual operation undergo multiple degradation stages, each characterized by its unique trend of degradation. The application of a singular stochastic process for RUL prediction falls short of achieving optimal performance. Consequently, this paper introduces a multi-stage Wiener process-based approach for the prediction of bearings' RUL. Initially, to tackle the challenge of imbalanced sample sizes across different degradation stages of bearings, an ensemble learning-based neural network (ELN) enhanced by ARIMA Residual Anomaly Detection (ARAD) for the identification of bearing degradation stages is proposed. Subsequently, taking into account the temporal, unit-to-unit, and nonlinear variabilities of the degradation process at each stage, a Wiener process-based multi-stage degradation model for bearings is developed. A method for parameter estimation and updating, utilizing Kalman filtering and Maximum Likelihood Estimation (K-M) is introduced. Finally, the proposed model undergoes validation using both simulated data and the XJTU-SY bearing dataset. Experimental results of three RUL predictions show that the proposed method leads the benchmark model with RMSE values of 3.61, 2.92 and 7.24 respectively, affirming that the proposed model can precisely classify equipment degradation stages and predict RUL with high accuracy and stability.

Integrating Learning-Driven Model Behavior and Data Representation for Enhanced Remaining Useful Life Prediction in Rotating Machinery

The increasing complexity of modern mechanical systems, especially rotating machinery, demands effective condition monitoring techniques, particularly deep learning, to predict potential failures in a timely manner and enable preventative maintenance strategies. Health monitoring data analysis, a widely used approach, faces challenges due to data randomness and interpretation difficulties, highlighting the importance of robust data quality analysis for reliable monitoring. This paper presents a two-part approach to address these challenges. The first part focuses on comprehensive data preprocessing using only feature scaling and selection via random forest (RF) algorithm, streamlining the process by minimizing human intervention while managing data complexity. The second part introduces a Recurrent Expansion Network (RexNet) composed of multiple layers built on recursive expansion theories from multi-model deep learning. Unlike traditional Rex architectures, this unified framework allows fine tuning of RexNet hyperparameters, simplifying their application. By combining data quality analysis with RexNet, this methodology explores multi-model behaviors and deeper interactions between dependent (e.g., health and condition indicators) and independent variables (e.g., Remaining Useful Life (RUL)), offering richer insights than conventional methods. Both RF and RexNet undergo hyperparameter optimization using Bayesian methods under variability reduction (i.e., standard deviation) of residuals, allowing the algorithms to reach optimal solutions and enabling fair comparisons with state-of-the-art approaches. Applied to high-speed bearings using a large wind turbine dataset, this approach achieves a coefficient of determination of 0.9504, enhancing RUL prediction. This allows for more precise maintenance scheduling from imperfect predictions, reducing downtime and operational costs while improving system reliability under varying conditions.

Switching Unscented Kalman Filters With Unknown Transition Probabilities for Remaining Useful Life Prediction of Bearings

Switching Unscented Kalman Filters With Unknown Transition Probabilities for Remaining Useful Life Prediction of Bearings

Reliability and fatigue life prediction of GFRP pipes by continuum damage mechanics-based damage elastic–plastic bridging model

Reliability and fatigue life prediction of GFRP pipes by continuum damage mechanics-based damage elastic–plastic bridging model

Bayesian Domain-Adversarial Regression Adaptation: A New Self-Driven Remaining Useful Life Prediction Method Across Different Machines With Uncertainty Quantification

Bayesian Domain-Adversarial Regression Adaptation: A New Self-Driven Remaining Useful Life Prediction Method Across Different Machines With Uncertainty Quantification

Prediction of rotational bending fatigue life of bearing steel based on damage mechanics

Based on the theory of continuous damage mechanics, this paper proposes a fatigue damage evolution model and numerical calculation method considering the influence of vacuum chemical heat treatment process, and studies the rotational bending fatigue damage analysis and life prediction method of M50NiL bearing steel. Firstly, the constitutive model, fatigue damage evolution equation and material parameter calibration method in the theoretical model are given. Secondly, based on the ABAQUS platform, the damage mechanics-finite element numerical calculation method of fatigue damage analysis considering the influence of hardened layer is realized by writing the UMAT subroutine; then, the rotational bending fatigue test of M50NiL bearing steel treated by vacuum chemical heat treatment of quenching and tempering, carburizing and carbonitriding is carried out, and the influence of heat treatment process on fatigue performance is analyzed; finally, based on the proposed fatigue damage model and numerical calculation method, the rotational bending fatigue life of M50NiL bearing steel is predicted, and the results are compared with the experimental results to verify the correctness of the proposed method.

Remaining Useful Life Prediction Method Based on Dual-Path Interaction Network with Multiscale Feature Fusion and Dynamic Weight Adaptation

In fields such as manufacturing and aerospace, remaining useful life (RUL) prediction estimates the failure time of high-value assets like industrial equipment and aircraft engines by analyzing time series data collected from various sensors, enabling more effective predictive maintenance. However, significant temporal diversity and operational complexity during equipment operation make it difficult for traditional single-scale, single-dimensional feature extraction methods to effectively capture complex temporal dependencies and multi-dimensional feature interactions. To address this issue, we propose a Dual-Path Interaction Network, integrating the Multiscale Temporal-Feature Convolution Fusion Module (MTF-CFM) and the Dynamic Weight Adaptation Module (DWAM). This approach adaptively extracts information across different temporal and feature scales, enabling effective interaction of multi-dimensional information. Using the Commercial Modular Aero-Propulsion System Simulation (C-MAPSS) dataset for comprehensive performance evaluation, our method achieved RMSE values of 0.0969, 0.1316, 0.086, and 0.1148; MAPE values of 9.72%, 14.51%, 8.04%, and 11.27%; and Score results of 59.93, 209.39, 67.56, and 215.35 across four different data categories. Furthermore, the MTF-CFM module demonstrated an average improvement of 7.12%, 10.62%, and 7.21% in RMSE, MAPE, and Score across multiple baseline models. These results validate the effectiveness and potential of the proposed model in improving the accuracy and robustness of RUL prediction.

Service life prediction model for biogenic acid corrosion: Advancement of life factor method for concrete sewer design

Current service life prediction models for concrete sewer design, including the widely used Life Factor Method (LFM), do not efficiently incorporate certain key corrosion contributing factors, such as material selection, concrete mix design, and sewer environment. This paper describes an advancement of the LFM model to cover these factors, broaden its usage, and improve concrete sewer design and applications. By integrating the first principles of the original LFM model, advanced knowledge from the literature, and results from field performance studies of various sewer concretes, a practical and relatively simple LFM model was advanced. With concrete mix design, oxide composition of all concrete ingredients, and sewer environmental and hydraulic conditions, the so-called ‘advanced LFM model’ could predict the corrosion rate of various concrete mixes. The model emphasised the influence of various modern binders, including calcium aluminate and calcium sulpho-aluminate cements, and various aggregate types including siliceous, calcareous, and alumina-based aggregates, when subjected to different sewer environments.

Implementation of an app-based time-temperature-indicator system for the real-time shelf life prediction in a pork sausage supply chain

Time-temperature-indicators (TTIs) combined with predictive modeling are helpful tools for avoiding the increasing amount of food waste and the associated waste of resources along the food supply chain. Successful implementation of these systems in practice is still absent due to missing digital technologies for real-time shelf life prediction. This study aimed to validate a novel app system developed for the digital read-out of OnVu™ TTIs and the shelf life prediction of perishable products along the raw pork sausage supply chain. Therefore, a kinetic shelf life model of raw pork sausage was developed based on microbial parameters. A dynamic TTI model was developed based on app-measured TTI data and validated on a laboratory scale to prepare for the pilot study. In the pilot study, the shelf life prediction of TTIs based on app measurements was validated under practical conditions. Results showed that the spoilage kinetics of raw pork sausage could, in general, be reflected by the OnVu™ TTI kinetics based on the app's color measurements. The pilot study showed that predicted and measured TTI color values by the app were in good agreement with accuracy factors of 1.02–1.03; however, slight differences in shelf lives revealed that the prediction model must be further improved by integrating more data. Although variances in the hourly range could be seen between the predicted shelf life based on TTI app measurements and the real shelf life of raw pork sausage, the study serves as a proof of concept for the general useability of the app for shelf life prediction because it showed that TTI and product kinetics were highly comparable. Further technical adjustments to the app and the adaptation of the charging time may further improve the shelf life prediction by the app along the raw pork sausage supply chain.

A dual-channel convolutional neural network with attention mechanism DC_EcaNet-6 for creep life prediction of notched components

Machine learning models offer novel possibilities for creep life prediction of materials and components at elevated temperatures. Present studies primarily focus on material-level creep life prediction, with limited reports on component-level analysis due to the complex stress states around structural discontinuities. Based on this, a dual-channel convolutional neural network with attention mechanism, DC_EcaNet-6, is proposed for creep life prediction of notched components, where the images of Mises stress and stress tri-axiality are employed as the input. The prediction results by the DC_EcaNet-6 model are compared with that by some deep learning models and the simplified method. Creep reliability assessment of notched components using the DC_EcaNet-6 model is conducted. The results indicate that the proposed model provides a more superior creep life prediction accuracy of notched components than other models mentioned. This model provides a potential tool for creep life prediction and reliability assessment of notched components.

An Analytical Approach for IGBT Life Prediction Using Successive Variational Mode Decomposition and Bidirectional Long Short-Term Memory Networks

The precise estimation of the operational lifespan of insulated gate bipolar transistors (IGBT) holds paramount significance for ensuring the efficient and uncompromised safety of industrial equipment. However, numerous methodologies and models currently employed for this purpose often fall short of delivering highly accurate predictions. The analytical approach that combines the Pattern Optimization Algorithm (POA) with Successive Variational Mode Decomposition (SVMD) and Bidirectional Long Short-term Memory (BiLSTM) network is introduced. Firstly, SVMD is employed as an unsupervised feature learning method to partition the data into intrinsic modal functions (IMFs), which are used to eliminate noise and preserve the essential signal. Secondly, the BiLSTM network is integrated for supervised learning purposes, enabling the prediction of the decomposed sequence. Additionally, the hyperparameters of BiLSTM and the penalty coefficients of SVMD are optimized utilizing the POA technique. Subsequently, the various modal functions are predicted utilizing the trained prediction model, and the individual mode predictions are subsequently aggregated to yield the model’s definitive final life prediction. Through case studies involving IGBT aging datasets, the optimal prediction model was formulated and its lifespan prediction capability was validated. The superiority of the proposed method is demonstrated by comparing it with benchmark models and other state-of-the-art methods.

Fatigue life modeling for a low alloy steel after long-term thermal aging

Fatigue failure of materials has a considerable impact on the safety of equipment in service. In this study, axially tensile and low cycle fatigue tests were conducted on a low alloy steel after accelerated thermal aged at 450 °C for 10,000 h. The experimental results indicate that the ultimate and yield strengths increase moderately, while the fatigue life of specimens experience a slight decrease in this circumstance. The fracture analysis demonstrates that the bainite breaking facilitates the fatigue crack initiation and propagation after thermal aging, which is accompanied by a decrease in plastic strain amplitude. Therefore, the plastic strain amplitude is considered as an indicator of thermal aging in fatigue life modeling for the low alloy steel. Finally, a novel life model that incorporates both aging time and temperature was proposed for rapid prediction of low cycle fatigue life. It is assumed that this model promotes reliable fatigue life prediction in low alloy steels under various thermal aging circumstances, as well as the extrapolation of fatigue performance of the material in service.

Remaining useful life prediction of lithium-ion batteries based on DBO[sbnd]CNN-DSformer

In order to improve the accuracy of predicting RUL of lithium-ion batteries, a lithium-ion battery RUL prediction method based on the DBOCNN-DSformer model is proposed. Firstly, the health characteristics of the battery are extracted and the local information of health features is mined using CNN. DSformer is utilized for global information, local information, and variable correlation learning of battery aging characteristics. The DBO is used to optimize the super-parameters of the CNN-DSformer model and build the DBOCNN-DSformer model. Finally, the battery aging data set was used for verification. The results show that DBOCNN-DSformer, which sets different prediction starting points, can extract sequence information from input data and establish long-term dependencies between sequences. The maximum average MRE error in the NASA data set was 0.05, the maximum average MAE was 0.018, and the maximum average AE error was within 5. The maximum average MRE error of the CALCE data set was 0.37, the maximum average MAE was 0.014, and the maximum average AE prediction error was within 10. Compared with LSTM, RNN, and Transformer models, it was found that DBOCNN-DSformer showed high prediction accuracy and good robustness.