Prognostics Research Articles

Remaining useful life prediction of nuclear reactor control rod drive mechanism based on dynamic temporal convolutional network

The control rod drive mechanism (CRDM) is a critical equipment of the nuclear reactor, and the prediction of its remaining useful life (RUL) is important for the efficient maintenance and ensuring the safe, reliable operation of nuclear power plants. In this paper, a novel framework for the RUL prediction of CRDM is proposed, which is a dynamic temporal convolution network (DTCN) based on dynamic activation function and attention mechanism. Firstly, the temporal convolution network (TCN) is used as the backbone of the prediction model, to extract the temporal dependence of the input data. Next, the dynamic activation function DReLU is integrated into the TCN, which can dynamically activate features and capture variable degradation information. Then, introducing the attention mechanism improves the influence of important high-level features extracted by the network on RUL prediction, thereby improving the efficiency of feature extraction in the network. Finally, the DTCN outputs the predicted RUL by performing non-linear mapping on the extracted features. The CRDM accelerated life test platform is established and a series of experiments are conducted using the collected CRDM full-life vibration dataset. The results demonstrated the performance and advantages of the proposed method.

Degradation Modeling and RUL Prediction of Hot Rolling Work Rolls Based on Improved Wiener Process.

Hot rolling work rolls are essential components in the hot rolling process. However, they are subjected to high temperatures, alternating stress, and wear under prolonged and complex working conditions. Due to these factors, the surface of the work rolls gradually degrades, which significantly impacts the quality of the final product. This paper presents an improved degradation model based on the Wiener process for predicting the remaining useful life (RUL) of hot rolling work rolls, addressing the critical need for accurate and reliable RUL estimation to optimize maintenance strategies and ensure operational efficiency in industrial settings. The proposed model integrates pulsed eddy current testing with VMD-Hilbert feature extraction and incorporates a Gaussian kernel into the standard Wiener process to effectively capture complex degradation paths. A Bayesian framework is employed for parameter estimation, enhancing the model's adaptability in real-time prediction scenarios. The experimental results validate the superiority of the proposed method, demonstrating reductions in RMSE by approximately 85.47% and 41.20% compared to the exponential Wiener process and the RVM model based on a Gaussian kernel, respectively, along with improvements in the coefficient of determination (CD) by 121% and 19.76%. Additionally, the model achieves reductions in MAE by 85.66% and 42.61%, confirming its enhanced predictive accuracy and robustness. Compared to other algorithms from the related literature, the proposed model consistently delivers higher prediction accuracy, with most RUL predictions falling within the 20% confidence interval. These findings highlight the model's potential as a reliable tool for real-time RUL prediction in industrial applications.

Asymmetric-Based Residual Shrinkage Encoder Bearing Health Index Construction and Remaining Life Prediction.

Predicting the remaining useful life (RUL) of bearings is crucial for maintaining the reliability and availability of mechanical systems. Constructing health indicators (HIs) is a fundamental step in the methodology for predicting the RUL of rolling bearings. Traditional HI construction often involves determining the degradation stage of the bearing by extracting time-frequency domain features from raw data using a priori knowledge and setting artificial thresholds; this approach does not fully utilize the vibration information in the bearing data. In order to address the above problems, this paper proposes an Asymmetric Residual Shrinkage Convolutional Autoencoder (ARSCAE) model. The asymmetric structure of the ARSCAE model is characterized by the soft thresholding of signal features in the encoder part to achieve noise reduction. The decoder part consists of convolutional and pooling layers for data reconstruction. This model can directly construct HIs from the original vibration signals collected, and comparisons with other models show that it constructs better HIs from the original vibration signals. Finally, experiments on the FEMTO dataset show that the results indicate that the HIS constructed by the ARSCAE model has better lifetime prediction capability compared to other methods.

Application of DBN-based KRLS method for RUL prediction of lithium-ion batteries

The Remaining Useful Life (RUL) of lithium batteries is vital for maintaining and safely operating the batteries, making precise RUL predictions highly significant. This paper introduces a method for predicting the RUL of lithium-ion batteries, utilizing a kernel adaptive filtering algorithm integrated with Deep Belief Networks (DBN). The method constructs a novel prediction model based on the Fixed-Budget Kernel Recursive Least Squares (FB-KRLS) algorithm. In this approach, the DBN extracts features from the original lithium battery data to reduce data complexity. The Square-root Cubature Kalman Filter (SCKF) is integrated with the FB-KRLS algorithm, employing a dual al-ternating learning strategy to improve the model's nonlinear fitting performance. The model was validated using NASA's lithium battery data, showing that the minimum val-ues for the MAPE, RMSE and MAE were 0.102%, 0.0016 and 0.0014, respectively. Therefore, the proposed method demonstrates potential for application in predicting the RUL of lithium-ion batteries.

Remaining useful life prediction of high-capacity lithium-ion batteries based on incremental capacity analysis and Gaussian kernel function optimization

Remaining useful life (RUL) is a key indicator for assessing the health status of lithium (Li)-ion batteries, and realizing accurate and reliable RUL prediction is crucial for the proper operation of battery systems. As high-capacity Li batteries have more complex chemical properties, most of the current RUL prediction methods rely mainly on a priori knowledge to make judgments. As a result, prediction accuracy is not high. In this study, we developed a health indicator-capacity (HI-C) dual Gaussian process regression (GPR) model based on incremental capacity analysis (ICA) and optimized its kernel function to achieve accurate RUL prediction for 280 Ah high-capacity Li batteries. Validation against the United States National Aeronautics and Space Administration’s four battery test datasets showed that the use of the HI-C dual GPR model resulted in a mean absolute percentage error and root mean square error of less than 0.02 and 0.04, respectively, for the four battery-rated capacity predictions. Additionally, this model achieved an absolute error of less than five battery failure turns. Compared with a single model, the HI-C dual GPR model not only had high accuracy but also solved the problem that the HI was not measurable in the actual battery operation, which made it more suitable for RUL prediction of Li batteries.

A closed-form continuous-depth neural-based hybrid difference features re-representation network for RUL prediction

Remaining Useful Life (RUL) prediction contributes to ensuring the reliability of mechanical systems and improving their maintenance plans. Recently, prediction methods based on deep learning have undergone rapid development. However, there are significant differences between multivariate monitoring sequences and utilizing a single model for feature extraction, which leads to reaching suboptimization. Additionally, focusing solely on a specific feature dimension usually imposes notable limitations on the model. This paper proposes a Closed-form Continuous-depth neural (CfC)-based hybrid difference Features Re-representation Network (CfC-F2RN) for RUL prediction. This method comprehensively utilizes hybrid difference features through two stages: initial feature representation and feature re-representation. In the initial feature representation stage, a Long Short-Term Memory (LSTM) is employed to capture the hidden state information of each time step in sequence data. Next, a novel attention mechanism is utilized to extract the hidden states and obtain a deep feature map. In the feature re-representation stage, the encoded features are re-represented using a decoder composed of the CfC model. Finally, the hidden state of the last LSTM unit in the encoder, serving as supplementary information, is combined with the decoded features to form a dual-latent feature representation of the mixed differential features. The prediction subnetwork is then adopted to accomplish RUL prediction. The superiority of CfC-F2RN is substantiated through benchmarking against established methods using the C-MAPSS dataset.

Remaining Useful Life of the Rolling Bearings Prediction Method Based on Transfer Learning Integrated with CNN-GRU-MHA

To accurately predict the remaining useful life (RUL) of rolling bearings under limited data and fluctuating load conditions, we propose a new method for constructing health indicators (HI) and a transfer learning prediction framework, which integrates Convolutional Neural Networks (CNN), Gated Recurrent Units (GRU), and Multi-head attention (MHA). Firstly, we combined Convolutional Neural Networks (CNN) with Gated Recurrent Units (GRU) to fully extract temporal and spatial features from vibration signals. Then, the Multi-head attention mechanism (MHA) was added for weighted processing to improve the expression ability of the model. Finally, a new method for constructing Health indicators (HIs) was proposed in which the noise reduction and normalized vibration signals were taken as a HI, the L1 regularization method was added to avoid overfitting, and the model-based transfer learning method was used to realize the RUL prediction of bearings under small samples and variable load conditions. Experiments were conducted using the PHM2012 dataset from the FEMTO-ST research institute and XJTU-SY dataset. Three sets of 12 migration experiments were conducted under three different operating conditions on the PHM2012 dataset. The results show that the average RMSE of the proposed method was 0.0443, indicating high prediction accuracy under variable loads and small sample conditions. Three different operating conditions and two sets of four migration experiments were conducted on the XJTU-SY dataset, and the results show that the average RMSE of the proposed method was 0.0693, verifying the good generalization of the model under variable load conditions. In summary, the proposed HI construction method and prediction framework can effectively reduce the differences between features, with high stability and good generalizability.

A novel performance prediction method for harmonic reducers based on the trend-enhanced Fisher’s discriminant ratio-Wiener process

Abstract Harmonic reducers, as core components of industrial robots, play a critical role in maintaining robot health. Performance degradation assessment (PDA) and remaining useful life (RUL) prediction are essential for ensuring the operational reliability of such robots. To address the multistage characteristics and stochastic uncertainties in the performance degradation of harmonic reducers, a performance prediction method based on Fisher’s discriminant ratio and Wiener process is proposed. Firstly, the health indicator (HI) construction is defined as an optimization problem based on Fisher’s discriminant ratio and trend monotonicity constraints. By leveraging the sum of the multiplication of frequency amplitudes and optimized weights, precise segmentation of the multistage degradation process over the entire lifecycle is achieved. Notably, the optimized weights can automatically identify the resonance frequency bands caused by damage. Subsequently, a nonlinear Wiener process model with a drift coefficient in the form of a power function is established. The probability density function (PDF) expression for RUL is derived based on the concept of first hit time (FHT). Additionally, a logarithmic likelihood function (Log-LF) for the unknown parameters in the degradation model is constructed. The experimental results indicate that the proposed method surpasses the other two Wiener process models in terms of root mean square error (RMSE), mean absolute percentage error (MAPE), and cumulative relative accuracy (CRA). This provides robust support for preventive maintenance decision-making for harmonic reducers.

Prediction model optimization of gas turbine remaining useful life based on transfer learning and simultaneous distillation pruning algorithm

For the application of deep learning (DL) models in the field of remaining useful life (RUL) prediction and predictive maintenance (PdM) of complex equipment, the insufficient training data and large model are the two major problems. To address these issues, a model training method based on transfer learning and a simultaneous distillation pruning algorithm were proposed. By introducing prior knowledge, three transfer learning modes are devised to reduce the demand of training data. Additionally, the simultaneous distillation pruning algorithm was devised to make the model lightweight, and an iterative pruning method was adopted to trim the large neural network model. By analyzing the performance of different transfer learning modes, the effectiveness of the proposed method can be demonstrated. The number of model parameters and the performance before and after pruning were compared. The results demonstrated that, without significant alterations to the prediction performance, the proposed model exhibited the capability to markedly reduce the number of model parameters. Based on the proposed methods, the challenges of insufficient data and efficiency encountered by DL models could be effectively addressed.

Sinkhorn divergence-based contrast domain adaptation for remaining useful life prediction of rolling bearings under multiple operating conditions

Under multiple operating conditions, the degradation characteristics of rolling bearings show diverse distributions. Domain adaptation (DA) achieves effective alignment between source and target domains by extracting domain-invariant features. However, in the prediction of remaining useful life (RUL) for bearings, numerous DA methods overlook mutual information from target-specific data and encounter potential challenges such as the vanishing gradient problem during the alignment of data distributions, leading to limited performance. To address these challenges, a novel method called Sinkhorn Divergence-based Contrast Domain Adaptation (SD_CDA) is proposed to predict RUL under multiple operating conditions. Firstly, an adversarial training framework is constructed to initially extract domain-invariant features. Subsequently, the cross-domain temporal mixup strategy is proposed for the data augment, which obtains positive samples to serve contrastive learning. Then self-supervised momentum contrast (MoCo) is employed to extract mutual information from target-specific data, preserving its specificity. Finally, Sinkhorn divergence is introduced to further align the fine-grained structure of the source domain and target domain, and enhance the transfer ability of the model. The experimental results demonstrate the superiority and effectiveness of the proposed method under multiple operating conditions.

A comprehensive framework for estimating the remaining useful life of Li-ion batteries under limited data conditions with no temporal identifier

The escalating applications of Lithium-ion (Li-ion) batteries in renewable energy and electric vehicles underscore the need for enhanced prognostics and health management systems to reduce the risk of sudden failures. Remaining useful life (RUL) determination is one of the most critical tasks in the field of battery prognostics nowadays. Even though statistical and machine learning (ML) methods have proven effective in research setups, many challenges prevent applying these prediction methods to real-life scenarios. These challenges include (1) scarcity of run-to-failure datasets with similar experimental conditions, (2) low data granularity when presented in capacity vs. discharge cycle pairs, and (3) lack of “temporal identifiers” in real-life scenarios. A temporal identifier is any label that provides knowledge about the current degradation state of a working battery. The research question developed for this study was, ‘Can the remaining useful life of a Li-ion battery having limited data without a temporal identifier be predicted?’ The specific aims were to estimate the temporal identifier of limited data and to predict the remaining useful life (RUL). An innovative framework incorporating reliability analysis and deep learning addresses these specific aims. Experimental data is used to test the framework's capabilities, limiting the training dataset to only three batteries and the testing dataset to a small sample (< 10 data points) of another battery. This new approach enabled the RUL prediction to achieve errors as low as ∼5 cycles and root mean square error of 6.24 cycles, outperforming other benchmark studies on Li-ion battery RUL prediction that use more battery degradation data without temporal identifier.

A self-supervised assisted label-efficient method for online remaining useful life prediction

Remaining Useful Life (RUL) prediction is crucial for safe and efficient production in industrial equipment. Conventional methods struggle in online environments, facing challenges like data distribution differences across working conditions, model cold start problem under new equipment, and lack of labels in online scenarios. To address these challenges, we propose a novel framework for online RUL prediction that leverages knowledge from self-supervised tasks via metric learning. First, a label-efficient self-supervised learning method incorporating soft triplet loss is proposed to transfer degradation knowledge. Second, we develop a two-stage method to distinguish different degradation stages of the equipment for better alignment of data distribution. Third, we introduce a pseudo-label dynamic self-updating strategy and a sample entropy-based data replay strategy, fine-tuning the offline model with newly acquired data. Our method outperforms state-of-the-art unsupervised approaches, reducing RMSE and MAPE by 16.02% and 8.40%, respectively, and achieves performance comparable to supervised methods.

Fisher-informed continual learning for remaining useful life prediction of machining tools under varying operating conditions

Accurate prediction of remaining useful life (RUL) of equipment has become an essential task in manufacturing. It not only helps prevent unexpected failures but also enables maximal utilization of available life, thus improving process efficiency. In practice, however, the use of multiple operating conditions that vary by time impedes efficient data-driven RUL prediction. Unlike conventional supervised learning setups, varying operating conditions generate heterogeneous data with time-varying generating distributions. Thus, existing approaches cannot be effectively applied due to increasing modeling and memory costs. One of the domains that suffer from this issue is machining, where RUL prediction of cutting tools is crucial for productivity. Considering realistic circumstances with varying operating conditions, this work proposes a method named Fisher-informed continual learning (FICL), which enables efficient tool RUL prediction that adaptively learns as conditions change without storing previous data and models. FICL uses Fisher information to improve generalization via sharpness-aware minimization and transfer knowledge between operating conditions through structural regularization. Experiments using datasets from real-world machining processes under five distinct operating conditions prove FICL’s efficacy, indicating its superior prediction performance to existing methods for all operating conditions. Particularly, FICL manifests the least catastrophic forgetting, implying it effectively retains informative knowledge from varying operating conditions.

An aero-engine remaining useful life prediction model based on clustering analysis and the improved GRU-TCN

Abstract Accurately predicting the remaining useful life (RUL) of engines is paramount for implementing effective preventive maintenance strategies, preventing injuries and fatalities caused by equipment failures, and significantly reducing routine repair and replacement costs. However, existing deep learning models often ignore the variable operating conditions in real engineering applications and do not sufficiently consider the interaction between time series and degradation laws, which directly leads to the inability to effectively extract to degradation feature extraction. To address this problem, this study developed a novel combined network model named CA-DRGRU-TTCN, aimed at accurately predicting the RUL of engines. Firstly, a Density-Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm is used to identify multiple operating conditions, and incorporate the recognition results into the model as additional new features. The first degradation time point is determined by JS divergence. Secondly, the deep connectivity of the residual Gated Recurrent Unit (GRU) module is designed to extract deeper degradation features, and an improved TMSE loss function based on the first degradation time point is applied to Temporal Convolutional Networks (TCN) to better capture the dependency between the time series and the real degradation degree of the engine. Finally, experiment results on the C-MAPSS dataset show that the proposed method achieves better performance compared to existing state-of-the-art methods.&amp;#xD;

Remaining Useful Life Estimation of Lithium-Ion Batteries Based on Small Sample Models

Accurate prediction of the Remaining Useful Life (RUL) of lithium-ion batteries is essential for enhancing energy management and extending the lifespan of batteries across various industries. However, the raw capacity data of these batteries is often noisy and exhibits complex nonlinear degradation patterns, especially due to capacity regeneration phenomena during operation, making precise RUL prediction a significant challenge. Although various deep learning-based methods have been proposed, their performance relies heavily on the availability of large datasets, and satisfactory prediction accuracy is often achievable only with extensive training samples. To overcome this limitation, we propose a novel method that integrates sequence decomposition algorithms with an optimized neural network. Specifically, the Complementary Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN) algorithm is employed to decompose the raw capacity data, effectively mitigating the noise from capacity regeneration. Subsequently, Particle Swarm Optimization (PSO) is used to fine-tune the hyperparameters of the Bidirectional Gated Recurrent Unit (BiGRU) model. The final BiGRU-based prediction model was extensively tested on eight lithium-ion battery datasets from NASA and CALCE, demonstrating robust generalization capability, even with limited data. The experimental results indicate that the CEEMDAN-PSO-BiGRU model can reliably and accurately predict the RUL and capacity of lithium-ion batteries, providing a promising and reliable method for RUL prediction in practical applications.

A Dual-Dimension Convolutional-Attention Module for Remaining Useful Life Prediction of Aeroengines

Remaining useful life (RUL) prediction of aeroengines not only enhances aviation safety and operational efficiency but also significantly lowers operational costs, offering substantial economic and social benefits to the aviation industry. Aiming at RUL prediction, this paper proposes a novel dual-dimension convolutional-attention (DDCA) mechanism. DDCA consists of two branches: one includes channel attention and spatial attention mechanisms, while the other applies these mechanisms to the inverted dimensions. Pooling and feature-wise pooling operations are employed to extract features from different dimensions of the input data. These branches operate in parallel to capture more complex temporal and spatial feature correlations in multivariate time series data. Subsequently, an end-to-end DDCA-TCN network is constructed by integrating DDCA with a temporal convolutional network (TCN) for RUL prediction. The proposed prediction model is evaluated using the C-MAPSS dataset and compared to several state-of-the-art RUL prediction models. The results show that the RMSE and SCORE metrics of DDCA-TCN decreased by at least 12.8% and 4.6%, respectively, compared to other models on the FD002 subset, and by at least 10.6% and 18.4%, respectively, on the FD004 subset, which demonstrates that the DDCA-TCN model exhibits excellent performance in RUL prediction, particularly under multiple operating conditions.

Intelligent fatigue damage tracking and prognostics of composite structures utilizing raw images via interpretable deep learning

In recent years, prognostics gained attention in various industries by optimizing maintenance, boosting operational efficiency, and preventing costly downtime. Central to prognostics is the Remaining Useful Life (RUL), representing the critical time before system failure. Deep learning advancements facilitate RUL forecasting by extracting features from diverse data formats such as time series, images, or sequences thereof, in one, two, or three dimensions, respectively. Yet, predicting RUL from image sequences often relies heavily on resource-intensive techniques like digital image correlation, complicating data acquisition. To address challenges with high-dimensional data and unreliable models, this study introduces ISTRUST, an innovative Transformer-based architecture. ISTRUST (Interpretable Spatiotemporal TRansformer for Understanding STructures) tackles the dual challenges posed by high-dimensional data and the black-box nature of existing models. Leveraging Transformers’ attention mechanism, ISTRUST breaks down the spatiotemporal domain, effectively realizing interpretable RUL predictions under uncertainty using only sparse raw image sequences as input. Evaluated on fatigue-loaded composite samples showcasing crack propagation, ISTRUST interprets the relation between cracks and RUL via the attention mechanism. The results substantiate its capacity to interpret and clarify instances in which predictions may exhibit variability in accuracy. Through the attention mechanism, a strong correlation between the model’s spatiotemporal focus and the RUL predictions is established, making it, to the best of our knowledge, the first model to provide interpretable stochastic RUL predictions directly from sequential images of this nature.

RUL forecasting for wind turbine predictive maintenance based on deep learning

Predictive maintenance (PdM) is increasingly pursued to reduce wind farm operation and maintenance costs by accurately predicting the remaining useful life (RUL) and strategically scheduling maintenance. However, the remoteness of wind farms often renders current methodologies ineffective, as they fail to provide a sufficiently reliable advance time window for maintenance planning, limiting PdM's practicality. This study introduces a novel deep learning (DL) methodology for future-RUL forecasting. By employing a multi-parametric attention-based DL approach that bypasses feature engineering, thereby minimizing the risk of human error, two models—ForeNet-2d and ForeNet-3d—are proposed. These models successfully forecast the RUL for seven multifaceted wind turbine (WT) failures with a 2-week forecast window. The most precise forecast deviated by only 10 minutes from the actual RUL, while the least accurate prediction deviated by 1.8 days, with most predictions being off by only a few hours. This methodology offers a substantial time frame to access remote WTs and perform necessary maintenance, thereby enabling the practical implementation of PdM.

A novel two-stage method via adversarial strategy for remaining useful life prediction of bearings under variable conditions

It is critical to accurately predict the remaining useful life (RUL) of rolling bearings to avoid severe accidents and financial losses in the industry. Nevertheless, accurately determining the initial prediction time (IPT) continues to pose a challenge, and significant differences in the data distribution of bearings under different operating conditions are frequently overlooked. To deal with these problems, we propose a novel two-stage method based on the adversarial strategy for RUL prediction of bearings under variable conditions. Firstly, we create reliable health indicators in an unsupervised manner by recording the coded characteristics of the bearing’s state of health. Secondly, an adaptive threshold method based on rate-of-change (ATMROC) is developed to perform accurate health state classification. Finally, we propose a RUL prediction network based on the attention depth-gated recurrent unit with domain invariance (DIADGRU) to handle the inconsistent distribution of degradation features under different operating conditions. Experiments of RUL prediction on PHM2012 and XITU-SY datasets are implemented, and the promising results validate the effectiveness of the proposed method.

A remaining useful life prediction method based on DATCN-PSOSEN for turbofan engines

Abstract Turbofan engine is a crucial operational component of aircraft. Remaining useful life (RUL) prediction is important for the stable and reliable operation of the turbofan engine. High-dimensional and large-capacity monitoring data of turbofan engines pose a considerable challenge to accurate RUL prediction. A novel approach based on dual-attention temporal convolutional network (DATCN) and particle swarm optimization with selective ensemble (PSOSEN) is proposed in this paper. In the first stage, DATCN is utilized to explore the internal correlations among various input features and different time steps in monitoring data, highlighting the degradation information from two dimensions. In the second stage, PSOSEN is developed to prune base models, excluding the ones with poor performance and assigning varying weights to the others, leading to selectively ensembled prediction results. The experimental results on the C-MAPSS aero-engine degradation dataset validate the effectiveness of the proposed DATCN-PSOSEN method and show improvements of RUL prediction accuracy by 14.2% compared with other methods.