ARTICLE | doi:10.20944/preprints202311.1705.v1
Subject: Engineering, Safety, Risk, Reliability And Quality Keywords: transformer; self-supervised learning; autoencoder; remaining useful life prediction; bidirectional LSTM; turbofan engine
Online: 27 November 2023 (13:22:46 CET)
Estimating the Remaining Useful Life (RUL) of aircraft engines holds a pivotal role in enhancing safety, optimizing operations, and promoting sustainability, thus being a crucial component of modern aviation management. Precise RUL predictions offer valuable insights into an engine’s condition, enabling informed decisions regarding maintenance and crew scheduling. In this context, we propose a novel RUL prediction approach in this paper, harnessing the power of Bi-directional LSTM and Transformer architectures, known for their success in sequence modeling, such as natural languages. We adopt the encoder part of the full Transformer as the backbone of our framework, integrating it with a self-supervised denoising autoencoder that utilizes Bidirectional LSTM for improved feature extraction. Within our framework, a sequence of multivariate time series sensor measurements serves as the input, initially processed by the Bidirectional LSTM autoencoder to extract essential features. Subsequently, these feature values are fed into our Transformer encoder backbone for RUL prediction. Notably, our approach simultaneously trains the autoencoder and Transformer encoder, different from the naive sequential training method. Through a series of numerical experiments carried out on the C-MAPSS datasets, we demonstrate that the efficacy of our proposed models either surpasses or stands on par with that of other existing methods.
ARTICLE | doi:10.20944/preprints201907.0112.v1
Subject: Computer Science And Mathematics, Computer Science Keywords: autoencoder; prognostics and health management; outlier modeling; remaining useful life; gear bearing
Online: 8 July 2019 (10:08:05 CEST)
In this paper, we introduce the Prognostics and Health Management of gear bearing system using autoencoder neural networks. Bearings and gears are the most common mechanical components in rotating machines, and their health conditions are of great concern in practice. This study presents an outlier modeling method for forecasting the gear bearing system failure using the health indicators constructed from mechanical signal processing and modeling. Outlier modeling aims to find patterns in data that are significantly different from what is defined as normal. In the unsupervised outlier modeling setting, prior labels about the anomalousness of data points are not available. In such cases, the most common techniques for scoring data points for outlyingness include distance-based methods density-based methods, and linear methods. The conventional outlier modeling methods have been used for a long time to detect anomalous observations in data. However, this paper shows that autoencoders are a very competitive technique compared to other existing methods. The developed method is demonstrated using the IMS bearing data from NASA Acoustics and Vibration Database.
ARTICLE | doi:10.20944/preprints202306.1480.v1
Subject: Engineering, Energy And Fuel Technology Keywords: Remaining useful life; Empirical mode decomposition; Bayesian optimization algorithm; Multi-kernel relevance vector machine; PEMFC
Online: 21 June 2023 (08:55:25 CEST)
To predict the remaining useful life (RUL) of proton exchange membrane fuel cell (PEMFC) in advance, a prediction method based on the voltage recovery model and Bayesian optimization of a multi-kernel relevance vector machine (MK-RVM) is proposed in this paper. First, the empirical mode decomposition (EMD) method was used to preprocess the data, and then the MK-RVM was used to train the model. Then, the Bayesian optimization algorithm was used to optimize the weight coefficient of the kernel function to complete the parameter update of the prediction model, and the voltage recovery model was added to the prediction model to realize the rapid and accurate prediction of the RUL of PEMFC. Finally, the method proposed in this paper was applied to the open data set of PEMFC provided by FCLAB, and the prediction accuracy of the RUL of PEMFC was obtained by 95.35%, which showed that the method had good gen-eralization ability and verified the accuracy of the prediction for the RUL of PEMFC.
ARTICLE | doi:10.20944/preprints202001.0054.v1
Subject: Engineering, Energy And Fuel Technology Keywords: lithium-ion (Li-ion) battery; remaining useful life (RUL); health indicator (HI); generalized regression neural network (GRNN); non-linear autoregressive (NAR)
Online: 7 January 2020 (09:17:28 CET)
The remaining capacity can only be measured with offline method. This brings great challenge for the online prediction of Li-ion battery’s RUL. A novel online prediction method for Li-ion battery’s RUL was proposed, which is based on multiple health indicators (HIs) and can be derived from the batteries’ historical operation data. Firstly, four indirect HIs were built according to the battery’s operation current, voltage and temperature data respectively. On that basis, a generalized regression neural network (GRNN) was developed to estimate the battery’s remaining capacity, and the non-linear autoregressive approach (NAR) was utilized to predict the battery’s RUL based on the estimated capacity value. Furthermore, to reduce the interference, twice wavelet denoising were performed with different thresholds. A case study is conducted with a NASA battery dataset to demonstrate the effectiveness of the method. The result shows that the proposed method can obtain Li-ion batteries’ RUL effectively.
ARTICLE | doi:10.20944/preprints202309.0676.v1
Subject: Engineering, Mechanical Engineering Keywords: remaining useful life; maximum correlation kurtosis deconvolution; multi-scale permutation entropy; long short-term memory
Online: 11 September 2023 (11:30:22 CEST)
The performance of bearings plays a pivotal role in determining the dependability and security of rotating machinery. In intricate systems demanding exceptional reliability and safety, the ability to accurately forecast fault occurrences during operation holds profound significance. Such predictions serve as invaluable guides for crafting well-considered reliability strategies and executing maintenance practices aimed at enhancing reliability. In order to ensure the reliability of bearing operation, this article investigates the application of three advanced techniques—Maximum Correlation Kurtosis Deconvolution (MCKD), Multi-Scale Permutation Entropy (MPE), and Long Short-Term Memory (LSTM) recurrent neural networks—for the prediction of the remaining useful life (RUL) of rolling bearings. Each technique's principles, methodologies, and applications are comprehensively reviewed, offering insights into their respective strengths and limitations. Case studies and experimental evaluations are presented to assess their performance in RUL prediction. Findings reveal that MCKD enhances fault signatures, MPE captures complexity, and LSTM excels in modeling temporal patterns. The root mean square error of the prediction results is 0.007. The fusion of these techniques offers a comprehensive approach to RUL prediction, leveraging their unique attributes for more accurate and reliable predictions.
ARTICLE | doi:10.20944/preprints202108.0272.v1
Subject: Engineering, Industrial And Manufacturing Engineering Keywords: Remaining Useful Life; Deep Neural Network; Convolutional Neural Network; Genetic Optimization; Neural Network Optimization; Support Vector Regression; Depth Maps; Normal Maps; 3D Point Clouds.
Online: 12 August 2021 (10:40:23 CEST)
In the current industrial landscape, increasingly pervaded by technological innovations, the adoption of optimized strategies for asset management is becoming a critical key success factor. Among the various strategies available, the “Prognostics and Health Management” strategy is able to support maintenance management decisions more accurately, through continuous monitoring of equipment health and “Remaining Useful Life” forecasting. In the present study, Convolutional Neural Network-based Deep Neural Network techniques are investigated for the Remaining Useful Life prediction of a punch tool, whose degradation is caused by working surface deformations during the machining process. Surface deformation is determined using a 3D scanning sensor capable of returning point clouds with micrometric accuracy during the operation of the punching machine, avoiding both downtime and human intervention. The 3D point clouds thus obtained are transformed into bidimensional image-type maps, i.e., maps of depths and normal vectors, to fully exploit the potential of convolutional neural networks for extracting features. Such maps are then processed by comparing 15 genetically optimized architectures with the transfer learning of 19 pre-trained models, using a classic machine learning approach, i.e., Support Vector Regression, as a benchmark. The achieved results clearly show that, in this specific case, optimized architectures provide performance far superior (MAPE=0.058) to that of transfer learning which, instead, remains at a lower or slightly higher level (MAPE=0.416) than Support Vector Regression (MAPE=0.857).
ARTICLE | doi:10.20944/preprints202308.0836.v1
Subject: Engineering, Mechanical Engineering Keywords: rolling bearings；remaining useful life；grey wolf optimization algorithm；neural network
Online: 10 August 2023 (10:49:04 CEST)
In order to assess the degradation state of rolling bearings and accurately grasp the remaining life information of rolling bearings, a bearing remaining life prediction method based on the grey wolf optimization algorithm to improve the BP neural network model is proposed. The method consists of two steps, firstly, using the craggy value and root mean square value to determine the first failure time of rolling bearings so as to approximate the input features, and secondly, using the Grey Wolf optimization algorithm to optimize the BP neural network to construct the degradation model of the bearings through machine learning. The method reduces the prediction error compared with conventional techniques, indicating that the method can effectively simulate the bearing degradation process and predict the remaining useful life (RUL) of the bearing.
Subject: Engineering, Civil Engineering Keywords: life extension; wind turbines; end-of-life issues; probabilistic modelling; economic optimization; fatigue; risk; remaining useful life
Online: 18 January 2021 (15:02:18 CET)
Reassessment of the fatigue life for wind turbines structural components is typically performed using deterministic methods with the same partial safety factors as used for the original design. However, in relation to life extension, the conditions are generally different from the assumptions used for calibration of partial safety factors; and using a deterministic assessment method with these partial safety factors might not lead to optimal decisions. In this paper, the deterministic assessment method is compared to probabilistic and risk-based approaches, and the economic feasibility is assessed for a case wind farm. Using the models also used for calibration of partial safety factors in IEC61400-1 ed. 4 it is found that the probabilistic assessment generally leads to longer additional fatigue life than the deterministic assessment method. The longer duration of the extended life can make life extension feasible in more situations. The risk-based model is applied to include the risk of failure directly in the economic feasibility assessment and it is found that the reliability can be much lower than the target for new turbines, without compromising the economic feasibility.
ARTICLE | doi:10.20944/preprints202112.0218.v1
Subject: Engineering, Mechanical Engineering Keywords: concrete; remote sensing; remaining life assessment; condition assessment
Online: 13 December 2021 (17:45:55 CET)
Concrete condition assessing penetrometers need to be able to distinguish between making contact with a hard (concrete) surface as opposed to a semi-solid (corroded concrete) surface. If a hard surface is mistaken for a soft surface, concrete corrosion may be over-estimated, with the potential for triggering unnecessary remediation works. Unfortunately, the variably-angled surface of a concrete pipe can cause the tip of a force-sensing tactile penetrometer to slip and thus to make this mistake. We investigated whether different shaped tips of a cylindrical penetrometer were better than others at maintaining contact with concrete and not slipping. We designed a range of simple symmetric tip shapes, controlled by a single superellipse parameter. We performed a finite element analysis of these parametric models in SolidWorks before machining in stainless steel. We tested our penetrometer tips on a concrete paver cut to four angles at 20∘ increments. The results indicate that penetrometers with a squircle-shaped steel tip (a=b=1,n=4) have the least slip, in the context of concrete condition assessment.
ARTICLE | doi:10.20944/preprints201910.0238.v1
Subject: Computer Science And Mathematics, Artificial Intelligence And Machine Learning Keywords: hybrid machine learning model; transportation infrastructure; flexible pavement; remaining service life prediction; pavement condition index; support vector regression; fruit fly optimization algorithm (foa); gene expression programming (gep); svr-foa
Online: 20 October 2019 (17:11:10 CEST)
Remaining service life (RSL) of pavement, as a sign of future pavement performance, has always received growing attention from pavement engineers. The RSL describes the time from the moment of pavement inspection until such a time when a major repair or reconstruction is required. The conventional approach to determining RSL involves using non-destructive tests. These tests, in addition to being costly, interfere with traffic flow and compromise users' safety. In this paper, surface distresses of pavement have been used to estimate the pavement’s RSL in order to eliminate the aforementioned problems and challenges. To implement the proposed theory, 105 flexible pavement segments were taken from Shahrood-Damghan Highway (Highway 44) in Iran. For each pavement segment, the type, severity, and extent of surface damage and pavement condition index (PCI) were determined. The pavement RSL was then estimated using non-destructive tests include Falling Weight Deflectometer (FWD) and Ground Penetrating Radar (GPR). After completing the dataset, the modeling was conducted to predict RSL using three techniques include Support Vector Regression (SVR), Support Vector Regression Optimized by Fruit Fly Optimization Algorithm (SVR-FOA), and Gene Expression Programming (GEP). All three techniques estimated the RSL of the pavement by selecting the PCI as input. The Correlation Coefficient (CC), Nash-Sutcliffe efﬁciency (NSE), Scattered Index (SI), and Willmott’s Index of agreement (WI) criteria were used to examine the performance of the three techniques adopted in this study. In the end, it was found that GEP with values of 0.874, 0.598, 0.601, and 0.807 for CC, SI, NSE, and WI criteria, respectively, had the highest accuracy in predicting the RSL of pavement.
ARTICLE | doi:10.20944/preprints202312.0384.v1
Subject: Engineering, Electrical And Electronic Engineering Keywords: battery; capacity; degradation; state of health; remaining useful life; neural networks; wavelets
Online: 6 December 2023 (10:25:37 CET)
Lithium-ion batteries are key elements in the development of electrical energy storage solutions. However, due to cycling, environmental and operating conditions, battery capacity tends to degrade over time. Capacity fade is a common indicator of battery state of health (SOH) because it is an indication of how capacity has been degraded. However, battery capacity cannot be measured directly, so there is an urgent need to develop methods for estimating battery capacity in real time. By analyzing the historical data of a battery in detail, it is possible to predict the future state of the battery and forecast its remaining useful life. This paper develops a real-time, simple and fast method to estimate the cycle capacity of a battery during the charge cycle using only data from a short period of each charge cycle. This proposal is attractive because it does not require data from the entire charge period, since batteries are rarely charged from zero to full. The proposed method allows simultaneous and accurate real-time prediction of the health and remaining useful life of the battery over its lifetime. The accuracy of the proposed method has been tested using experimental data from several lithium-ion batteries with different cathode chemistries under various test conditions.
ARTICLE | doi:10.20944/preprints202005.0386.v1
Subject: Public Health And Healthcare, Public Health And Health Services Keywords: remaining useful life; c-mapss; extreme learning machine; prognostic and health management; neural networks
Online: 24 May 2020 (16:24:08 CEST)
This work can be considered as a first step of designing a future competitive data-driven approach for remaining useful life prediction of aircraft engines. The proposed approach is an ensemble of serially connected extreme learning machines. The results of prediction of the first networks are scaled and fed to the next networks as an additive features to the original inputs. This feature mapping allows increasing the correlation of training inputs with their targets by holding new prior knowledge about the probable behavior of the target function. The proposed approach is evaluated under remaining useful estimation using a set of “time-varying” data retrieved from the public dataset C-MAPSS (Commercial Modular Aero Propulsion System Simulation) provided by NASA. The prediction performances are compared to basic extreme learning machine and proved the effectiveness of the proposed methodology.
ARTICLE | doi:10.20944/preprints202011.0591.v2
Subject: Engineering, Automotive Engineering Keywords: health indicator; performance degradation assessment; deep learning; vibration monitoring; bearing; remaining useful life; digital twin
Online: 11 December 2020 (11:59:20 CET)
Estimating the remaining useful life (RUL) of components is a crucial task to enhance the reliability, safety, productivity, and to reduce maintenance cost. In general, predicting the RUL of a component includes constructing a health indicator ( ) to infer the current condition of the component, and modelling the degradation process, to estimate the future behavior. Although many signal processing and data-driven based methods were proposed to construct the , most of the existing methods are based on manual feature extraction techniques, and need the prior knowledge of experts, or rely on a large amount of failure data. Therefore, in this study, a new data-driven method based on the convolutional autoencoder (CAE) is presented to construct the . For this purpose, the continuous wavelet transform (CWT) technique is used to convert the raw acquired vibrational signals into a two-dimensional image; then, the CAE model is trained by the healthy operation dataset. Finally, the Mahalanobis Distance (MD) between the healthy and failure stages is measured as the . The proposed method is tested on a benchmark bearing dataset and compared with several other traditional construction models. Experimental results indicate that the constructed exhibits a monotonically increasing degradation trend and has a good performance to detect incipient faults.
DATA DESCRIPTOR | doi:10.20944/preprints202301.0473.v1
Subject: Computer Science And Mathematics, Applied Mathematics Keywords: Data generator; dataset; deep learning; health index; machine learning; prognosis and health management; remaining useful life
Online: 26 January 2023 (08:37:30 CET)
This paper presents PrognosEase; a software that provides an easier way to produce different types of run-to-failure data mimicking real-world conditions to simplify prognosis studies in terms of data collection and improvement in ML degradation modelling process. Different types of degradation types made available to meet different types of applications. Besides, some primary ML tests were performed to ensure that complexity patterns of real systems could be observed in the training/testing predictions attitude. This paper also presents the impacts, limitations and potential improvements of the data generator.
ARTICLE | doi:10.20944/preprints202303.0003.v1
Subject: Computer Science And Mathematics, Artificial Intelligence And Machine Learning Keywords: XAI; Interpretability; Predictive Maintenance; Prognostics and Health Management; Remaining Useful Life; Deep Learning; Convolutional Neural Network; Comparative Analysis; XAI Evaluation and Accountability
Online: 1 March 2023 (02:31:28 CET)
The aim of Predictive Maintenance, within the field of Prognostics and Health Management (PHM), is to identify and anticipate potential issues in the equipment before these become critical. The main challenge to be addressed is to assess the amount of time a piece of equipment will function effectively before it fails, which is known as Remaining Useful Life (RUL). Deep Learning (DL) models, such as Deep Convolutional Neural Networks (DCNN) and Long Short-Term Memory (LSTM) networks, have been widely adopted to address the task, with great success. However, it is well known that this kind of black box models are opaque decision systems, and it may be hard to explain its outputs to stakeholders (experts in the industrial equipment). Due to the large number of parameters that determine the behavior of these complex models, understanding the reasoning behind the predictions is challenging. This work presents a critical and comparative revision on a number of XAI methods applied on time series regression model for PM. The aim is to explore XAI methods within time series regression, which have been less studied than those for time series classification. The model used during the experimentation is a DCNN trained to predict the RUL of an aircraft engine. The methods are reviewed and compared using a set of metrics that quantifies a number of desirable properties that any XAI method should fulfill. The results show that GRAD-CAM is the most robust method, and that the best layer is not the bottom one, as is commonly seen within the context of Image Processing.