Denoising Autoencoder Research Articles

Tool Condition Monitoring Model Based on DAE–SVR

Cutting tools are executive components in metal processing, and tool wear directly affects the quality of the workpiece and processing efficiency; monitoring the change in its state is crucial to avoid accidents and ensure the safety of workers. The traditional monitoring model cannot compress a large amount of cutting data effectively, failing to obtain reliable feature data, and there are some defects in generalization ability and monitoring accuracy. For this purpose, this article takes milling cutters as the research object, and it integrates signals from force sensors, vibration sensors, and acoustic emission sensors, combining the advantages of the denoising autoencoder (DAE) model in data compression and the high monitoring accuracy of the support vector regression (SVR) model, to establish a tool wear monitoring model based on DAE–SVR. The results show that compared with traditional DAE and SVR models in multiple datasets, the maximum improvement in monitoring performance (MAE) is 43.58%.

CSPDAE: Colorization via SuperPixel Downsampler Denoising AutoEncoder towards enhanced color structural clarity

Based on the various structural textures, sizes, and shapes of the gray image, the SuperPixel (SP)-based methods convert the rigid pixels into adaptable image patches that share common features. As a result, the SP-based colorization technique not only enhances the color appearance but also preserves the topological integrity of the images. Despite the effectiveness of neural network-based colorization methods, their integration with SP segmentation has traditionally been complex and cumbersome. To address this issue, we propose the Colorization SP Downsampler Denoising AutoEncoder (CSPDAE), where the SP downsampler integrates SP segmentation directly into the network, eliminating the need for any prior input. The SP downsampler addresses the challenges of computational complexity and small region clustering, which are the main obstacles preventing Transformer architectures from being applied to pixel-level segmentation, through the use of SP cross-attention and aggregated positional embedding (APE). Furthermore, we have incorporated a Color Weight (CW) loss, based on the CIEDE2000 color difference formula, to ensure balanced pixel sampling and to improve the precision of detailed color representation. The experimental results confirm the effectiveness of our method, demonstrating its capacity to produce colors with greater structural accuracy and visually appealing details.

Multi-Layered Model Approach (MLMA) For Heart Vein Blockage Detection Using Gated Recurrent Unit (GRU) and Denoising Autoencoders (DAES)

Cardiovascular diseases (CVD) continue to be the leading cause of death worldwide, and sophisticated diagnostic techniques must be developed to detect and treat them early. This study analyzed vein blockages in the heart and proposed an ensemble model for the early diagnosis of heart problems. Venous blockages can be found using non-invasive imaging techniques, frequently suggesting cardiovascular abnormalities. Using a Coronary Computed Tomography Angiography (CCTA) dataset that includes normal and abnormal cardiac cases, the pre-trained 3D convolutional neural network (3D-CNN) model can learn complex patterns linked to early-stage cardiac problems. To improve the accuracy and dependability of the detection process, the proposed ensemble model incorporates several algorithms, including the segmentation-based threshold method, feature extraction-based form descriptors, and the vessel tracking algorithm (VTA). The ensemble model consisted of several base models, each focusing on a distinct area of vein blockage investigation. Feature extraction techniques extract pertinent information from medical imaging data, such as angiograms and vascular scans. The ensemble model, which uses the combined decision-making ability of several algorithms, including shape descriptors, is fed to the retrieved features. The proposed ensemble model performed better than the individual models and conventional diagnostic techniques regarding early detection accuracy. Moreover, analysis of the feature importance scores improves the interpretability of the model and sheds light on the crucial markers of cardiac anomalies. The suggested ensemble model has a significant potential clinical impact because early identification of cardiac problems enables prompt intervention and better patient outcomes. Finally, the proposed method combines the multi-layer approach with gated recurrent units (GRU) and a Denoising Autoencoder (DAE) to find abnormalities in angiography images. The effectiveness of the suggested method demonstrates a high rate of disease diagnosis and accuracy for specified angiography images.

Application of machine learning techniques for warfarin dosage prediction: a case study on the MIMIC-III dataset.

Warfarin, a commonly prescribed anticoagulant, poses significant dosing challenges due to its narrow therapeutic range and high variability in patient responses. This study applies advanced machine learning techniques to improve the accuracy of international normalized ratio (INR) predictions using the MIMIC-III dataset, addressing the critical issue of missing data. By leveraging dimensionality reduction methods such as principal component analysis (PCA) and t-distributed stochastic neighbor embedding (t-SNE), and advanced imputation techniques including denoising autoencoders (DAE) and generative adversarial networks (GAN), we achieved significant improvements in predictive accuracy. The integration of these methods substantially reduced prediction errors compared to traditional approaches. This research demonstrates the potential of machine learning (ML) models to provide more personalized and precise dosing strategies that reduce the risks of adverse drug events. Our method could integrate into clinical workflows to enhance anticoagulation therapy in cases of missing data, with potential applications in other complex medical treatments.

Neural network-based method for structured light phase-shifted fringe denoising and phase-to-height calibrations

Abstract The monocular structured light measurement system is widely applied across various fields due to its simple structure, low cost, and extensive measurement range. However, measurement accuracy can be affected by noise, non-linear intensity variations, variations in object surface reflectance, and calibration precision, leading to unstable or inaccurate results. Traditional filtering methods are limited in effectively addressing complex noise and non-linear issues, and phase-to-height calibration often depends on high-precision motion platforms, which increases system costs, complexity, and calibration uncertainty. To address these challenges, this paper proposes a neural network-based approach for monocular structured light measurement. An iterative denoising algorithm based on a denoising autoencoder (DAE) is developed for phase-shift fringe images. By optimizing the iteration count, denoised phase-shift fringes are regenerated to de noise to reduce image noise and enhance phase computation accuracy effectively. Additionally, leveraging the high-precision absolute phase calculated from the denoised phase-shift fringe images, a multilayer feedforward neural network (FNN) algorithm is proposed for absolute phase-to-height calibration. This method directly maps the phase-to-height relationship, integrating both intrinsic and extrinsic camera parameters to achieve high-precision calibration without requiring a high-precision motion platform. Consequently, it significantly mitigates errors related to the motion platform and reduces operational errors during calibration. The DAE enhances the quality of phase images, providing more precise input for the FNN calibration and further improving measurement accuracy. Experimental results demonstrate that the proposed method achieves effective 3D reconstruction from low-quality phase-shift fringes and maintains robust performance when measuring objects with varying reflectances.

Noise Reduction in CWRU Data Using DAE and Classification with ViT

With the Fourth Industrial Revolution unfolding worldwide, technologies including the Internet of Things, sensors, and artificial intelligence are undergoing rapid development. These technological advancements have played a significant role in the dramatic growth of the predictive maintenance market for mechanical equipment, prompting active research on noise removal techniques and classification algorithms for the accurate determination of the causes of equipment failure. In this study, time series data were preprocessed using the denoising autoencoder technique, a deep learning-based noise removal method, to improve the accuracy of failure classification from mechanical equipment data. To convert the preprocessed time series data into frequency components, the short-time Fourier transform technique was employed. The fault types of mechanical equipment were classified using the vision transformer (ViT) technique, a deep learning technique that has been actively used in recent image analysis research. Additionally, the classification performance of the ViT-based technique for vibration time series data was comparatively validated against existing classification algorithms. The accuracy of failure classification was the highest when the data, preprocessed using a Denoising Autoencoder (DAE), were classified by a Vision Transformer (ViT).

Short-term Traffic Flow Prediction：A Method of MEA-LSTM Model Based on Chaotic Characteristics Analysis

To effectively improve the accuracy of short-term traffic flow prediction, an improved Long Short-Term Memory (LSTM) method is proposed using the Mind Evolution Algorithm (MEA). Firstly, to address the issues of abnormal and missing traffic flow data, a Neighborhood Stacked Denoising AutoEncoder (NSDAE) is used for data repair. Then, the maximum Lyapunov exponent is used to determine the chaotic characteristics. Meanwhile, based on Bayesian estimation theory, the features of three-parameter sequences are fused in high-dimensional space using phase space reconstruction technique to obtain reconstructed multi-parameter fused traffic flow data. Finally, by taking advantage of the fact that MEA can divide the data into several subpopulations for optimal search separately, a prediction model based on MEA to improve LSTM is proposed. The results show that compared to the other two traditional data restoration methods, the NSDAE has higher accuracy, with the lowest average values of RMSE, MAE, and MAPE. Through the phase space reconstruction technique, the feature fusion of three parameters of traffic flow is realized in high-dimensional space, which makes up for the insufficiency of a single time-series data that cannot comprehensively levy the characteristics of traffic flow. The MEA-LSTM model outperforms the LSTM model in terms of prediction accuracy, computational efficiency, and generalization ability, and its RMSE, MEA, and MAPE are reduced by 24.3%, 28.9%, and 30.1%, respectively.

Fatigue life prognosis of composite structures using a transferable deep reinforcement learning-based approach

Accurately predicting the remaining useful life (RUL) of Carbon Fiber Reinforced Polymer (CFRP) structures under fatigue loading is crucial for enhancing safety and minimizing maintenance costs, especially in industries like aerospace and automotive. However, the complex physical properties of CFRP, combined with the scarcity of real-world damage-condition data, make this task extremely challenging. To address these issues, we propose a novel deep reinforcement learning (DRL)-based prognostic method. Our approach integrates Denoising Autoencoder (DAE) and Transformer architectures to construct a powerful DRL Policy Network, capable of extracting high-quality features from X-ray records to capture the subtle progression of damage in CFRP structures. Additionally, we employ advanced data augmentation techniques to overcome the limitations of small datasets and introduce transfer learning to extend the model’s generalization capabilities across different CFRP structures. By pre-training on diverse CFRP datasets, our model achieves highly accurate RUL predictions for new designs, even with minimal labeled data from the target structure. Experimental results demonstrate that our method significantly outperforms current state-of-the-art (SOTA) techniques, offering a scalable, efficient, and practical solution for the real-world monitoring and prognostics of CFRP structures, with broad potential for industrial applications.

HARNESSING BLOCKCHAIN WITH ENSEMBLE DEEP LEARNING-BASED DISTRIBUTED DOS ATTACK DETECTION IN IOT-ASSISTED SECURE CONSUMER ELECTRONICS SYSTEMS

Consumer electronics (CE) and the Internet of Things (IoTs) are transforming daily routines by integrating smart technology into household gadgets. IoT allows devices to link and communicate from the Internet with better functions, remote control, and automation of various complex systems simulation platforms. The quick progress in IoT technology has continuously driven the progress of further connected and intelligent CEs, shaping more smart cities and homes. Blockchain (BC) technology is emerging as a promising technology offering immutable distributed ledgers that improve the security and integrity of data. However, even with BC resilience, the IoT ecosystem remains vulnerable to Distributed Denial of Service (DDoS) attacks. In contrast, the malicious actor overwhelms the network with traffic, disrupting services and compromising device functionality. Incorporating BC with IoT infrastructure presents groundbreaking techniques to alleviate these threats. IoT networks can better detect and respond to DDoS attacks in real time by leveraging BC cryptographic techniques and decentralized consensus mechanisms, which safeguard against disruptions and enhance resilience. There must be a reliable mechanism of recognition based on adequate techniques to detect and identify whether these attacks have happened or not in the system. Artificial intelligence (A) is the most common technique that uses machine learning (ML) and deep learning (DL) to recognize cyber threats. This research presents a new Blockchain with Ensemble Deep Learning-based Distributed DoS Attack Detection (BCEDL-DDoSD) approach in the IoT platform. The primary intention of the BCEDL-DDoSD approach is to leverage BC with a DL-based attack recognition process in the IoT platform. BC technology is utilized to enable a secure data transmission process. In the BCEDL-DDoSD approach, Z-score normalization is initially employed to measure the input data. Besides, the selection of features takes place using the Fractal Wombat optimization algorithm (WOA). For attack recognition, the BCDL-DDoSD technique applies an ensemble of three models, namely denoising autoencoder (DAE), gated recurrent unit (GRU), and long short-term memory (LSTM). Lastly, an orca predator algorithm (OPA)-based hyperparameter tuning procedure has been implemented to select the parameter value of DL models. A sequence of simulations is made on the benchmark database to authorize the performance of the BCDL-DDoSD approach. The simulation results showed that the BCDL-DDoSD approach performs better than other DL techniques.

Nonlinear Dynamic Process Monitoring Based on Discriminative Denoising Autoencoder and Canonical Variate Analysis

Modern industrial processes are characterized by increasing complexity, often exhibiting pronounced dynamic behaviors and significant nonlinearity. Addressing these dynamic and nonlinear characteristics is essential for effective process monitoring. However, many existing methods for process monitoring and fault diagnosis are insufficient in handling these challenges. In this article, we present a novel process monitoring approach, CVA-DisDAE, which integrates an improved Denoising Autoencoder (DAE) with Canonical Variate Analysis (CVA) to address the challenges posed by dynamic behaviors and nonlinear relationships in industrial processes. First, CVA is employed to reduce data dimensionality and minimize information redundancy by maximizing correlations between past and future observations, thereby effectively capturing process dynamics. Following this, we introduce a discriminative DAE model (DisDAE) designed to serve as a semi-supervised denoising autoencoder for precise feature extraction. This is achieved by incorporating both between-class separability and within-class variability into the traditional DAE framework. The key distinction between the proposed DisDAE and the conventional DAE lies in the integration of a linear discriminant analysis (LDA) penalty into the DAE’s loss function, resulting in extracted features that are more conducive to fault classification. Finally, we validate the effectiveness of the proposed semi-supervised dynamic process monitoring approach through its application to the Tennessee Eastman benchmark process, demonstrating its superior performance.

Demodulating Optical Wireless Communication of FBG Sensing with Turbulence-Caused Noise by Stacked Denoising Autoencoders and the Deep Belief Network

Free-space optics communication (FSO) can be used as a transmission medium for fiber optic sensing signals to make fiber optic sensing easier to implement; however, interference with the sensing signals caused by the optical turbulence and scattering of airborne particles in the FSO path is a potential problem. This work aims to deep denoise sensed signals from fiber Bragg grating (FBG) sensors based on FSO link transmission using advanced denoising deep learning techniques, such as stacked denoising autoencoders (SDAE). Furthermore, it will demodulate the sensed wavelength of FBGs by applying the deep belief network (DBN) technique. This is the first time the real FBG sensing experiment has utilized the actual noise interference caused by the environmental turbulence from an FSO link rather than adding noise through numerical processing. Consequently, the spectrum of the FBG sensors is clearly modulated by the noise and the issue with peak power variation. This complicates the determination of the center wavelengths of multiple stacked FBG spectra, requiring the use of machine learning techniques to predict these wavelengths. The results indicate that SDAE is efficient in denoising from the FBG spectrum, and DBN is effective in demodulating the central wavelength of the overlapped FBG spectrum. Thus, it is beneficial to implement an FSO link-based FBG sensing system in adverse weather conditions or atmospheric turbulence.

Self-supervised learning for effective denoising of flow fields

In this study, we proposed an efficient approach based on a deep learning (DL) denoising autoencoder (DAE) model for denoising noisy flow fields. The DAE operates on a self-learning principle and does not require clean data as training labels. Furthermore, investigations into the denoising mechanism of the DAE revealed that its bottleneck structure with a compact latent space enhances denoising efficacy. Meanwhile, we also developed a deep multiscale DAE for denoising turbulent flow fields. Furthermore, we used conventional noise filters to denoise the flow fields and performed a comparative analysis with the results from the DL method. The effectiveness of the proposed DL models was evaluated using direct numerical simulation data of laminar flow around a square cylinder and turbulent channel flow data at various Reynolds numbers. For every case, synthetic noise was augmented in the data. A separate experiment used particle-image velocimetry data of laminar flow around a square cylinder containing real noise to test DAE denoising performance. Instantaneous contours and flow statistical results were used to verify the alignment between the denoised data and ground truth. The findings confirmed that the proposed method could effectively denoise noisy flow data, including turbulent flow scenarios. Furthermore, the proposed method exhibited excellent generalization, efficiently denoising noise with various types and intensities.

Computer-Assisted Online Learning of English Oral Pronunciation Based on DAE End-to-End Recurrent Neural Networks

With the development of globalization, learning a second language has received increasing attention from people. To improve English oral proficiency, a computer-aided online learning system for English oral pronunciation is studied. A denoising autoencoder is integrated into the system to create a simplified end-to-end recurrent neural network for pronunciation detection and diagnosis based on deep learning. The study first collected and preprocessed oral pronunciation data of English learners, including enhancing speech signals and reducing noise. Next, an RNN model with Long Short-Term Memory (LSTM) as the core was constructed to capture time series characteristics in pronunciation. And use DAE to extract features and reduce the influence of background noise to enhance the recognition of pronunciation features. At the same time, the study utilized web crawler technology to collect a large amount of oral pronunciation data from non-native English learners, and constructed an English oral corpus containing pronunciation errors. And in order to simulate real situations, white noise and pink noise were artificially added to the corpus in the study, and they were divided into training and testing sets in a ratio of 60% to 40%. The results showed that the classification accuracy of the system in the training and testing sets under white noise environment was 78.97% and 94.01%, respectively, and the classification accuracy in the pink noise environment was 76.19% and 94.03%, respectively. The system’s error detection accuracy in vowel and consonant pronunciation detection is 88.91% and 91.68%, respectively, and the error correction accuracy in vowel and consonant pronunciation detection is 90.67% and 91.96%, respectively. In summary, the research on computer-aided online learning of English oral pronunciation based on Denoising Auto Encoders end-to-end recurrent neural networks has effectively improved learning efficiency.

Deep feature fusion with computer vision driven fall detection approach for enhanced assisted living safety

Assisted living facilities cater to the demands of the elderly population, providing assistance and support with day-to-day activities. Fall detection is fundamental to ensuring their well-being and safety. Falls are frequent among older persons and might cause severe injuries and complications. Incorporating computer vision techniques into assisted living environments is revolutionary for these issues. By leveraging cameras and complicated approaches, a computer vision (CV) system can monitor residents’ movements continuously and identify any potential fall events in real time. CV, driven by deep learning (DL) techniques, allows continuous surveillance of people through cameras, investigating complicated visual information to detect potential fall risks or any instances of falls quickly. This system can learn from many visual data by leveraging DL, improving its capability to identify falls while minimalizing false alarms precisely. Incorporating CV and DL enhances the efficiency and reliability of fall detection and allows proactive intervention, considerably decreasing response times in emergencies. This study introduces a new Deep Feature Fusion with Computer Vision for Fall Detection and Classification (DFFCV-FDC) technique. The primary purpose of the DFFCV-FDC approach is to employ the CV concept for detecting fall events. Accordingly, the DFFCV-FDC approach uses the Gaussian filtering (GF) approach for noise eradication. Besides, a deep feature fusion process comprising MobileNet, DenseNet, and ResNet models is involved. To improve the performance of the DFFCV-FDC technique, improved pelican optimization algorithm (IPOA) based hyperparameter selection is performed. Finally, the detection of falls is identified using the denoising autoencoder (DAE) model. The performance analysis of the DFFCV-FDC methodology was examined on the benchmark fall database. A widespread comparative study reported the supremacy of the DFFCV-FDC approach with existing techniques.

Ultrasonic guided Wave Signal Denoising Method Based on one-dimensional convolution and Full connection Model Fusion Denoising autoencoder

Abstract Ultrasonic guided wave (UGW) detection is widely used in pipeline monitoring but faces challenges from weak flaw echo signals within the detection data, making weak UGW signals difficult to recognize. It is essential to denoise the UGW detection signals to identify weak echo signals. This paper proposes an improved denoising autoencoder (DAE) based on the fusion of one-dimensional convolution neural network (1DCNN) and full connection (FC). The model expands the amount of training data by adding noise in batches and uses 1DCNN to enhance the ability of extracting UGW signal features. The model was validated using Several numerical simulation signals. Numerical simulation results show that the signal-to-noise ratio (SNR) of the UGW signals can be improved from -20 dB to 8 dB; it has a strong improved SNR, and the mean square error is greatly reduced while maintaining the original phase almost unchanged. The improved DAE method has significant advantages in denoising effect, and it can effectively reduce the noise of the UGW detection signal and realize the identification of small defects of the simulation pipeline.

RUL Prediction of Rolling Bearings Based on Multi-Information Fusion and Autoencoder Modeling

As an important part of industrial equipment, the safe and stable operation of rolling bearings is an important guarantee for the performance of mechanical equipment. Aiming at the problem that it is difficult to characterize the running state of rolling bearings, this paper mainly analyzes and processes the vibration signals of rolling bearings, extracts and fuses multi-information entropy, and monitors the running state of rolling bearings and predicts the remaining useful life prediction (RUL) through test verification. Firstly, in view of the difficulty in characterizing the bearings running state characteristics, a rolling bearings running state monitoring method based on multi-information entropy fusion and denoising autoencoder (DAE) was proposed to extract the multi-entropy index features of vibration signals to improve the accuracy of feature extraction, and to solve the problem of not obvious information representation of a single feature indicator and missing information in the feature screening process. Secondly, in view of the problems of low prediction accuracy and poor robustness and generalization in traditional RUL models, a rolling bearings RUL model combining convolutional autoencoder (CAE) and bidirectional long short-term memory network (BiLSTM) was proposed. The introduction of convolution operation made CAE have the feature of weight sharing, reducing the complexity of the model. Finally, the XJTU-SY data set was used to verify the constructed model. The results show that the condition monitoring model established in this paper can accurately evaluate the running state of the rolling bearing and accurately locate the failure time. At the same time, the residual life prediction model can realize the residual life prediction of most data sets, and has good accuracy and robustness.

A comparison of machine learning methods for recovering noisy and missing 4D flow MRI data.

Experimental blood flow measurement techniques are invaluable for a better understanding of cardiovascular disease formation, progression, and treatment. One of the emerging methods is time-resolved three-dimensional phase-contrast magnetic resonance imaging (4D flow MRI), which enables noninvasive time-dependent velocity measurements within large vessels. However, several limitations hinder the usability of 4D flow MRI and other experimental methods for quantitative hemodynamics analysis. These mainly include measurement noise, corrupt or missing data, low spatiotemporal resolution, and other artifacts. Traditional filtering is routinely applied for denoising experimental blood flow data without any detailed discussion on why it is preferred over other methods. In this study, filtering is compared to different singular value decomposition (SVD)-based machine learning and autoencoder-type deep learning methods for denoising and filling in missing data (imputation). An artificially corrupted and voxelized computational fluid dynamics (CFD) simulation as well as in vitro 4D flow MRI data are used to test the methods. SVD-based algorithms achieve excellent results for the idealized case but severely struggle when applied to in vitro data. The autoencoders are shown to be versatile and applicable to all investigated cases. For denoising, the in vitro 4D flow MRI data, the denoising autoencoder (DAE), and the Noise2Noise (N2N) autoencoder produced better reconstructions than filtering both qualitatively and quantitatively. Deep learning methods such as N2N can result in noise-free velocity fields even though they did not use clean data during training. This work presents one of the first comprehensive assessments and comparisons of various classical and modern machine-learning methods for enhancing corrupt cardiovascular flow data in diseased arteries for both synthetic and experimental test cases.

Modality-Aware Discriminative Fusion Network for Integrated Analysis of Brain Imaging Genomics.

Mild cognitive impairment (MCI) represents an early stage of Alzheimer's disease (AD), characterized by subtle clinical symptoms that pose challenges for accurate diagnosis. The quest for the identification of MCI individuals has highlighted the importance of comprehending the underlying mechanisms of disease causation. Integrated analysis of brain imaging and genomics offers a promising avenue for predicting MCI risk before clinical symptom onset. However, most existing methods face challenges in: 1) mining the brain network-specific topological structure and addressing the single nucleotide polymorphisms (SNPs)-related noise contamination and 2) extracting the discriminative properties of brain imaging genomics, resulting in limited accuracy for MCI diagnosis. To this end, a modality-aware discriminative fusion network (MA-DFN) is proposed to integrate the complementary information from brain imaging genomics to diagnose MCI. Specifically, we first design two modality-specific feature extraction modules: the graph convolutional network with edge-augmented self-attention module (GCN-EASA) and the deep adversarial denoising autoencoder module (DAD-AE), to capture the topological structure of brain networks and the intrinsic distribution of SNPs. Subsequently, a discriminative-enhanced fusion network with correlation regularization module (DFN-CorrReg) is employed to enhance inter-modal consistency and between-class discrimination in brain imaging and genomics. Compared to other state-of-the-art approaches, MA-DFN not only exhibits superior performance in stratifying cognitive normal (CN) and MCI individuals but also identifies disease-related brain regions and risk SNPs locus, which hold potential as putative biomarkers for MCI diagnosis.

Acoustic signal adversarial augmentation for pressure pipeline leakage detection

Pressure pipelines are prone to leakage under harsh working condition for a long time, and the leakage detection reaches unsatisfactory performance due to influence of background noise and insufficient sample for acoustic signals. Therefore, the acoustic signals adversarial augmentation method is proposed for pressure pipeline leakage detection based on noise reduction and sample generation. By deeply connecting with generative adversarial network (GAN), denoising autoencoder (DAE) and residual network (ResNet), the adversarial denoising and generation model (ADGM) is established to reduce the noise of acoustic signal. In addition, the trained DAE of ADGM is applied to augment the acoustic samples, thereby completing adversarial augmentation of acoustic signal, which is significant for pressure pipeline leakage detection. Besides, the pipeline leakage experiment is implemented to validate the proposed method on noise reduction and sample generation, which can reach pressure pipeline detection accuracy of 93.02% based on augmented acoustic signal. Further, the effectiveness and superiority of the proposed method are tested by ablation experiments and comparative methods.

Robust remaining useful life prediction of lithium-ion battery with convolutional denoising autoencoder

Using lithium-ion (Li-ion) batteries exceeding their useful lifetime may be dangerous for users, and hence, developing an accurate prediction system for batteries that remain useful for life is necessary. Many deep learning models, such as gated recurrent units and long short-term memory (LSTM), have been proposed for that purpose and have shown good results. However, their performance when dealing with noisy data degrades significantly. This may hamper their implementations for the real world since battery data are prone to noise. In this paper, we develop a robust prediction model in a noisy environment for predicting the remaining useful life (RUL) of Li-ion batteries. We propose a denoising autoencoder (DAE) utilized to remove noise from the data. The DAE is built with convolutional layers instead of traditional feed-forward networks here. We combine DAE with LSTM as the predictor. The proposed framework is evaluated using artificially corrupted battery data provided by National Aeronautics and Space Administration (NASA). The results reveal that our proposed method improves robustness when data contain various types of noise. A comparative study using the traditional approach has also been conducted. Our evaluation shows that convolutional layers are more effective than the traditional approach and that the original composition of the DAE was built using traditional feed-forward networks. DAE with convolutional layers has the best average performance with MSE of 0.61 and is the most consistent model.