Continuous Wavelet Transform Research Articles (Page 1)

ABSTRACT Low-temperature charging induces lithium plating and stress accumulation, posing severe safety challenges for lithium-ion batteries (LIBs). However, the mechanisms underlying their internal damage evolution and failure mode transitions remain unclear. This study employs acoustic emission (AE) technology to develop an innovative hybrid classification model integrating K-means clustering, linear classification, and Gaussian kernel support vector machines. This approach enables adaptive recognition and dynamic tracking of multiple damage modes in LIBs. Results indicate that AE signals during charging exhibit distinctive dual-burst waveform characteristics. Pearson correlation analysis reveals that two burst signals share similar waveform features, originating from correlated waveforms of the same damage event. Classification results reveal that damage modes evolve from tensile-dominated patterns in the early charging stage to shear and mixed modes in the later stage. Furthermore, the coupled effects of low temperature and high charging rates significantly accelerate the accumulation of shear and mixed damage. Furthermore, continuous wavelet transform (CWT) analysis revealed a time-frequency evolution pattern where AE signals transitioned from high-frequency short-duration to low-frequency long-duration signals, aligning with the transformation of damage modes. This study established a multiscale acoustic emission analysis framework integrating hybrid learning classification and time-frequency analysis, providing novel insights and technical support for elucidating low-temperature failure mechanisms and enabling early warning in lithium-ion batteries.

Abstract Time–frequency analysis (TFA) has proven to be a powerful technique for analyzing non-stationary signals. However, in practical applications such as mechanical equipment monitoring, the collected vibration signals often exhibit strong non-stationarity and are heavily contaminated by noise. These challenges significantly limit the accuracy and robustness of conventional TFA methods. To address this issue, this paper proposes a local maximum synchroextracting wavelet transform (LMSEWT), a novel enhancement of the synchroextracting transform framework based on the continuous wavelet transform. LMSEWT introduces an adaptive mechanism into the time-frequency (TF) reassignment process by leveraging the multi-scale characteristics of wavelets and a local maximum criterion to refine instantaneous frequency estimation. This approach not only improves the resolution and concentration of TF representations but also enhances noise robustness. The theoretical foundation and implementation strategy of LMSEWT are detailed in this study. Its effectiveness is further demonstrated through application to fault diagnosis of variable-speed machinery, where it is benchmarked against several traditional TFA methods. Experimental results confirm the superior performance and practical value of the proposed method.

This paper addresses the long-standing question of understanding the origin and evolution of low-frequency unsteadiness interactions associated with shock waves impinging on a turbulent boundary layer in transonic flow (Mach: 1.1 to 1.3). To that end, high-speed experiments in a blowdown open-channel wind tunnel have been performed across a convergent–divergent nozzle for different expansion ratios (PR = 1.44, 1.6, and 1.81). Quantitative evaluation of the underlying spectral energy content has been obtained by processing time-resolved pressure transducer data and Schlieren images using the following spectral analysis methods: Fast Fourier Transform (FFT), Continuous Wavelet Transform (CWT), as well as coherence and time-lag evaluations. The images demonstrated the presence of increased normal shock-wave impact for PR = 1.44, whereas the latter were linked with increased oblique λ-foot impact. Hence, significant disparities associated with the overall stability, location, and amplitude of the shock waves, as well as quantitative assertions related to spectral energy segregation, have been inferred. A subsequent detailed spectral analysis revealed the presence of multiple discrete frequency peaks (magnitude and frequency of the peaks increasing with PR), with the lower peaks linked with large-scale shock-wave interactions and higher peaks associated with shear-layer instabilities and turbulence. Wavelet transform using the Morlet function illustrates the presence of varying intermittency, modulation in the temporal and frequency scales for different spectral events, and a pseudo-periodic spectral energy pulsation alternating between two frequency-specific events. Spectral analysis of the pixel densities related to different regions, called spatial FFT, highlights the increased influence of the feedback mechanism and coupled turbulence interactions for higher PR. Collation of the subsequent coherence analysis with the previous results underscores that lower PR is linked with shock-separation dynamics being tightly coupled, whereas at higher PR values, global instabilities, vortex shedding, and high-frequency shear-layer effects govern the overall interactions, redistributing the spectral energy across a wider spectral range. Complementing these experiments, time-resolved numerical simulations based on a transient 3D RANS framework were performed. The simulations successfully reproduced the main features of the shock motion, including the downstream migration of the mean position, the reduction in oscillation amplitude with increasing PR, and the division of the spectra into distinct frequency regions. This confirms that the adopted 3D RANS approach provides a suitable predictive framework for capturing the essential unsteady dynamics of shock–boundary layer interactions across both temporal and spatial scales. This novel combination of synchronized Schlieren imaging with pressure transducer data, followed by application of advanced spectral analysis techniques, FFT, CWT, spatial FFT, coherence analysis, and numerical evaluations, linked image-derived propagation and coherence results directly to wall pressure dynamics, providing critical insights into how PR variation governs the spectral energy content and shock-wave oscillation behavior for nozzles. Thus, for low PR flows dominated by normal shock structure, global instability of the separation zone governs the overall oscillations, whereas higher PR, linked with dominant λ-foot structure, demonstrates increased feedback from the shear-layer oscillations, separation region breathing, as well as global instabilities. It is envisaged that epistemic understanding related to the spectral dynamics of low-frequency oscillations at different PR values derived from this study could be useful for future nozzle design modifications aimed at achieving optimal nozzle performance. The study could further assist the implementation of appropriate flow control strategies to alleviate these instabilities and improve thrust performance.

ABSTRACT Soil organic carbon (SOC) content played a vital role in stabilizing oasis ecosystems and regulating carbon sequestration in arid lakeside regions. However, accurate estimation of SOC content using visible‐near‐infrared (Vis–NIR) spectroscopy was often hindered by spectral redundancy and high dimensionality. This study introduced a novel approach by integrating wavelet analysis with machine learning to improve SOC content estimation accuracy in the lakeside oasis of Bosten Lake, Xinjiang. A total of 82 topsoil (0–20 cm) samples were collected, and their SOC content and corresponding Vis–NIR spectra were measured. The hyperspectral data were processed using continuous wavelet transform (CWT) and discrete wavelet transform (DWT). Key spectral features were selected by three algorithms—successive projections algorithm (SPA), Boruta, and competitive adaptive reweighted sampling (CARS)—and used to develop SOC content estimation models based on partial least squares regression (PLSR), back propagation neural network (BPNN), and random forest (RF). Results indicated that CWT outperformed DWT in noise reduction, especially at low decomposition scales (1–5), with a 19.21% improvement. The best CWT‐based model yielded a 23.20% increase in residual prediction deviation (RPD) over the best DWT‐based model. Feature selection further enhanced model accuracy, improving the determination coefficient ( R 2 ) by up to 49.04% and RPD by 58.23%. Among the algorithms, CARS provided the highest improvement, followed by SPA and Boruta. Thus, the combination of CWT‐1‐CARS and the RF algorithm showed the strongest nonlinear modeling performance. The RF (CWT‐1‐CARS) configuration achieved calibration metrics of R 2 = 0.79, root mean square error (RMSE) = 2.57, and RPD = 2.23 to outperform the original spectral models, with an improvement of 63.3% over PLSR (RPD = 1.84) and BPNN (RPD = 1.91). The spatial interpolation analysis showed 91.3% consistency with field‐measured SOC content values, validating the model's practical reliability. The most sensitive spectral response bands for SOC content were primarily located in the visible range (401–504 nm) and the near‐infrared range (1638–2369 nm). This study established a robust technical foundation for accurate estimation of SOC content, for precise ecological monitoring, and sustainable management of arid, lakeside oases.

Reliable fault diagnosis of milling machines is essential for maintaining operational stability and cost-effective maintenance; however, it remains challenging due to limited labeled data and the highly non-stationary nature of acoustic emission (AE) signals. This study introduces an optimized Few-Shot Learning framework (FSL) that integrates time–frequency analysis with attention-guided representation learning and distribution-aware classification for data-efficient fault detection. The framework converts AE signals into Continuous Wavelet Transform (CWT) scalograms, which are processed using a self-attention-enhanced ResNet-50 backbone to capture both local texture features and long-range dependencies in the signal. Adaptive prototype computation with learnable importance weighting refines class representations, while Mahalanobis distance-based matching ensures robust alignment between query and prototype embeddings under limited sample conditions. To further strengthen discriminability, contrastive loss with hard negative mining enforces compact intra-class clustering and clear inter-class separation. Comprehensive experiments under 7-way 5-shot settings and 5-fold stratified cross-validation demonstrate consistent and reliable performance, achieving a mean accuracy of 98.86% ± 0.97% (95% CI: [98.01%, 99.71%]). Additional evaluations across multiple spindle speeds (660 rpm and 1440 rpm) confirm that the model generalizes effectively under varying operating conditions. Grad-CAM++ activation maps further illustrate that the network focuses on physically meaningful fault-related regions, enhancing interpretability. The results verify that the proposed framework achieves robust, scalable, and interpretable fault diagnosis using minimal labeled data, offering a practical solution for predictive maintenance in modern intelligent manufacturing environments.

EEG-based cerebral pattern analysis for neurological disorder detection via hybrid machine and deep learning approaches.

To address the limitations of single-modal approaches in bearing fault diagnosis under complex operating conditions, this study proposes SCBM-Net—a novel deep learning model based on a dual-channel multimodal fusion architecture. The model innovatively combines Continuous Wavelet Transform (CWT) and Variational Mode Decomposition (VMD) to extract complementary features from time–frequency images and temporal signals, respectively. Specifically, the first channel employs a Swin Transformer to effectively model both local and global representations of CWT-based images through a hierarchical window-based attention mechanism. The second channel adopts a CNN-BiGRU-Attention network to dynamically capture temporal dependencies from intrinsic mode functions decomposed by VMD. Features from both channels are deeply fused using a Multimodal Compact Bilinear Pooling (MCB) module, enhancing fault feature representation and overall model robustness. Experimental results on the CWRU dataset show that SCBM-Net achieves an accuracy of 99.83% under clean conditions. Even under a few-shot learning setting with only 60 training samples per class, the model still maintains a high recognition accuracy of 98.64%, demonstrating strong generalization in low-data scenarios. On an imbalanced dataset, SCBM-Net exhibits stable performance for both majority and minority classes, achieving an average accuracy of 97.33%. In a generalization test on the SEU bearing dataset, the model achieves an accuracy of 98.33%, further validating its cross-platform and cross-domain robustness and transferability. Moreover, under severe noise interference at − 10 dB, SCBM-Net retains a fault recognition accuracy of 80.67%, outperforming comparable models and demonstrating excellent noise robustness and practical applicability.

Covert visual attention decoding from EEG signals is a key challenge in cognitive neuroscience and brain–computer interface applications. Traditional approaches often rely on manual feature extraction and handcrafted pipelines, which limit scalability and generalization. In this study, we propose a deep learning-based framework that leverages time-frequency representations, specifically Continuous Wavelet Transform (CWT), to enable end-to-end classification of covert attention states without manual feature engineering. EEG data were recorded from ten healthy participants performing spatial and feature-based attention tasks. Among the tested models, ShallowConvNet achieved 100% accuracy in binary classification and over 90% in four-class conditions. EEGNet also performed competitively, exceeding 97% and 88% accuracy in two- and four-class scenarios, respectively. These findings demonstrate that integrating CWT with deep neural architectures significantly enhances decoding performance compared to conventional raw-signal approaches, offering a scalable and efficient solution for real-time attention monitoring.

The use of power generator sets (3.5kVA – 5.5kVA) for domestic and commercial backup supply has become a mainstay in Nigeria due to the unstable electricity from the grid. To keep these backup supplies running, traditional diagnostic approaches that are reliant on manual inspections and physical measurements have been adopted. These are often time-consuming, reactive, and unsuitable for real-time monitoring. To address these challenges, a machine learning approach is presented by performing a comparative analysis of MobileNet and AlexNet convolutional neural networks for automated audio-based fault diagnosis in 5kVA generators. Fault signatures are obtained from acoustic data recorded from 25 generator units under five operational states—Normal, Caburetter, Exhaust, Valve, and Plug faults. Mel-Frequency Cepstral Coefficients (MFCC), Continuous Wavelet Transform (CWT), and Short-Time Fourier Transform (STFT) were employed to transform the raw audio signals into two-dimensional spectrograms that contain both temporal and spectral fault signatures. Using transfer learning, these spectrograms were utilized as input features to train versions of MobileNet and AlexNet, which were pre-trained on ImageNet weights. Their performance was evaluated using accuracy, precision, recall, F1-score, and ROC-AUC metrics. Results obtained from the evaluation metrics show that MobileNet significantly outperformed AlexNet across all feature transformations (MFCC, CWT, and SSTFT). It achieved a peak accuracy of 92% and an AUC of 0.99 with STFT spectrograms. In contrast, AlexNet achieved lower accuracies (54–59%), indicating lower discriminative power. The class-wise ROC-AUC analyses confirmed that MobileNet achieved near-perfect classification, particularly in distinguishing between Normal and any of the fault conditions, while AlexNet struggled with subtle classes, such as Plug and Valve faults. These findings indicate that STFT is the most discriminative spectrogram and MobileNet is the best-performing diagnostic framework. This makes it suitable for deployment in resource-constrained environments and edge devices. This research contributes to the advancement of intelligent, real-time condition monitoring of domestic generator sets, thereby reducing downtime and enhancing energy reliability in off-grid contexts.

As a seismic isolation device, laminated rubber bearings are widely used in bridges. With increasing service life and under the effects of vehicular vibrations, earthquakes, and construction quality issues, these bearings frequently develop void damage, which poses risks to structural safety. However, manual inspection suffers from untimely detection, heavy workload, limited accuracy, and safety hazards. To rapidly and accurately detect the degree of void damage in bearings, this study proposes an innovative detection method combining the active sensing method, continuous wavelet transform (CWT), and a fine-tuned YOLOv5s model. A total of 2064 sets of detection signals from laminated rubber bearings in different void degrees were obtained using the active sensing method, and CWT was applied to convert the one-dimensional signals into two-dimensional time–frequency images. Subsequently, the pre-trained YOLOv5s model was fine-tuned to optimize its adaptability for void degree detection. Building on this foundation, a high-accuracy detection model was established. Across the training, validation, and test sets, all four metrics are identical, with accuracy = 100% and precision = recall = F1-score = 1.0. This indicates that the model’s predictions perfectly match the true labels. Compared with models trained on one-dimensional data or on two-dimensional time–frequency images, the proposed model exhibits a significant improvement in predictive performance. The proposed method shows strong potential for void damage detection in bridge bearings, enabling timely and high-accuracy diagnosis and thereby contributing to the safety of bridge structures.

Emotion recognition from EEG signals has emerged as a pivotal area of research, driven by its transformative potential in healthcare, brain-computer interfaces, and affective computing systems. However, the intrinsic complexity, non-linearity, and susceptibility to noise in EEG data present significant challenges to accurate emotional state classification. This study proposes a robust and interpretable hybrid deep learning model for EEG-based emotion recognition. The architecture integrates Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM) networks, and attention mechanisms, together with advanced signal processing techniques such as Continuous Wavelet Transform (CWT) and Power Spectral Density (PSD). This integrated approach facilitates the extraction of comprehensive spatial, temporal, and spectral features from EEG signals, enhancing the model’s ability to capture intricate patterns associated with emotional states. Experimental evaluations on the SEED-IV dataset, encompassing four emotional categories—Neutral, Happy, Sad, and Fear—demonstrated the model’s exceptional performance, achieving a macro-average F1-score of 93% and an area under the ROC curve (AUC) of 0.94. These results validate the model’s effectiveness in accurately distinguishing complex emotional patterns, even under noisy conditions and inter-class ambiguities. Overall, this research advances the domain of EEG-based emotion recognition by introducing a high-performing, interpretable framework suitable for real-world applications while laying the foundation for future developments in adaptive neurofeedback systems and emotion-aware brain-computer interfaces.

In recent years, the powerful feature extraction capabilities of deep learning have attracted widespread attention in the field of bearing fault diagnosis. To address the limitations of single-modal and single-channel feature extraction methods, which often result in incomplete information representation and difficulty in obtaining high-quality fault features, this paper proposes a dual-channel parallel multimodal feature fusion model for bearing fault diagnosis. In this method, the one-dimensional vibration signals are first transformed into two-dimensional time-frequency representations using continuous wavelet transform (CWT). Subsequently, both the one-dimensional vibration signals and the two-dimensional time-frequency representations are fed simultaneously into the dual-branch parallel model. Within this architecture, the first branch employs a combination of a one-dimensional convolutional neural network (1DCNN) and a bidirectional gated recurrent unit (BiGRU) to extract temporal features from the one-dimensional vibration signals. The second branch utilizes a dilated convolutional to capture spatial time–frequency information from the CWT-derived two-dimensional time–frequency representations. The features extracted by both branches were are input into the feature fusion layer. Furthermore, to leverage fault features more comprehensively, a channel attention mechanism is embedded after the feature fusion layer. This enables the network to focus more effectively on salient features across channels while suppressing interference from redundant features, thereby enhancing the performance and accuracy of the dual-branch network. Finally, the fused fault features are passed to a softmax classifier for fault classification. Experimental results demonstrate that the proposed method achieved an average accuracy of 99.50% on the Case Western Reserve University (CWRU) bearing dataset and 97.33% on the Southeast University (SEU) bearing dataset. These results confirm that the suggested model effectively improves fault diagnosis accuracy and exhibits strong generalization capability.

Parkinson's disease (PD) is a prevalent and complex neurodegenerative disorder, with early diagnosis playing a critical role in timely treatment and management. Handwriting dynamics has emerged as a promising biomarker for early detection of PD, yet current diagnostic methods often lack precision and robustness. This study introduces a novel multimodal deep learning-based decision support system to enhance PD diagnosis. Our approach leverages static and dynamic features of handwriting data by combining images of handwritten drawings with fused time-frequency representations of grip pressure, axial pressure, tilt, and accelerometer signals from the y- and z-axes recorded during handwriting. The time-frequency transformations employ Short-Time Fourier Transform (STFT) and Continuous Wavelet Transform (CWT) to generate spectrograms and scalograms. Results demonstrate that fusing STFT spectrograms achieves an accuracy of 85.41%, which improves to 97.92% when integrated into the multimodal CNN model. Similarly, fusing CWT scalograms achieves 92.08% accuracy, further enhanced to 96.66% with the multimodal approach. These findings highlight that fused time-frequency representations yield successful results for PD diagnosis. Furthermore, the CWT-based approach demonstrates superior performance compared to STFT. Finally, integrating fused time-frequency images with visualizations further improves the accuracy rates. We incorporate the Gradient-weighted Class Activation Mapping++(Grad-CAM++) eXplainable Artificial Intelligence (XAI) method to ensure interpretability, highlighting attention regions within the fused STFT and CWT images. These attention regions effectively differentiate between healthy controls (HC) and PD patients. Although the model achieved promising results on the NewHandPD dataset, further external validation on diverse and multi-center datasets is required to confirm its generalizability and clinical applicability. The findings underscore the potential of integrating handwriting-based static and dynamic features for high-precision PD diagnosis, offering a robust and explainable framework for clinical decision-making.

In the contemporary energy landscape, the increasing demand for electricity and the inherent uncertainties associated with the integration of renewable resources have rendered the accurate and reliable forecasting of short- and long-term demand imperative. Energy demand forecasting, fundamentally a time series problem, can be inherently complex, nonlinear, and multi-scale. Therefore, interest in artificial intelligence–based methods that provide high performance for short- and long-term forecasting, rather than traditional methods, has increased in order to solve these problems. In this study, a hybrid artificial intelligence model based on LSTM, GRU, and Random Forest, utilizing a distinct mechanism to address these types of problems, is proposed. The Multi-Scale Sliding Window (MSSW) approach was utilized for the model’s input data to capture the dynamics of the time series at different scales. The optimization of windows was conducted using the Continuous Wavelet Transform (CWT) method to determine the optimal window sizes within the MSSW structure in a data-driven manner. Experimental studies on Panama’s real energy demand data from 2015 to 2020 show that the CWT-aided MSSW-hybrid model forecasts better with lower error rates (0.007 MAE, 0.009 RMSE, 1.051% MAPE) than single models and manually determined window sizes. The results of the study demonstrate the importance of hybrid structures and window optimization in energy demand forecasting.

ABSTRACT Additive manufacturing (AM) allows the fabrication of complex, high-performance components, but these components are prone to porosity defects, which reduce their mechanical integrity and fatigue resistance. To enable real-time detection and enhance applicability to rough surfaces, this study proposes an intelligent porosity evaluation method for AM components based on grating laser ultrasonics. Specifically, the surface acoustic waves (SAWs) generated by a grating laser source interact with internal porosities, and the resulting time- and frequency-domain ultrasonic signals are converted into two-dimensional images by the Gramian angular field (GAF) algorithm. These images are subsequently processed using a Residual neural network (ResNet) model for porosity prediction. Firstly, through numerical simulations, the effects of porosity and surface roughness on both time-domain and frequency-domain signals are systematically analysed. Based on these results, various input feature representations (one-dimensional signals, continuous wavelet transform (CWT) images, and GAF images) are integrated with the ResNet architecture to perform porosity prediction on simulated and experimental datasets. After analysis, the GAF representation consistently achieves superior prediction performance under both smooth and rough surface conditions, demonstrating its strong generalisation ability and robustness. This validates the effectiveness and feasibility of the proposed method, providing a novel and reliable approach for accurate porosity assessment of AM components.

Power quality disturbances (PQDs) can significantly affect the reliability of electrical power systems, leading to potential equipment damage and operational inefficiencies. Accurate classification of these disturbances is essential for ensuring continuous and reliable service. The study proposes a deep transfer learning (TL) approach for PQD classification. In the proposed work, various single and multiple PQD signals pertaining to 15 different classes have been generated using mathematical models of PQDs adhering to the guidelines of the IEEE 1159 and IEC 61000-4-30 standards. Time-domain PQD signals are first converted into 2D color images using continuous wavelet transform (CWT). These images are then used to re-train modified pre-trained models such as GoogleNet, SqueezeNet, ResNet-18, and ShuffleNet on synthetic PQD data. Various single and combined PQDs are then classified using trained models. Moreover, the performance of the trained models is evaluated with the PQD signals containing noise of various signal-to-noise ratios (SNR), as well as PQD signals collected from the experimental setup in the laboratory. The results signify that the GoogleNet model exhibits consistent performance for classifying PQDs under various conditions, achieving classification accuracy of 99.8% for synthetic noiseless signals, 98.87% for signals with 20 dB SNR, and 98.89% for signals acquired through experimental setup. Furthermore, the trained models were tested using real-world PQD signals, comprising 26 sag and 42 impulse signals. The GoogleNet model achieved the highest classification accuracy, correctly identifying 23 sag and 34 impulse events, thereby demonstrating real-world applicability and robustness of the proposed approach.

An experimental study investigated pulsating air flow in circular pipes using acoustic excitation, with a focus on pressure drop-based flow resistance. Time-averaged pressure drop (ΔPta) measurements were analyzed by varying the Womersley number (Wo = 10–75), the pulsation amplitude (A = 20–50%), pipe diameter (d = 15–25 mm), and the Reynolds number (Re = 5000–8000). A peak increase in ΔPta of up to 260% was observed for Wo = 61, while a reduction of up to 77% was recorded for Wo > 66 at A = 35%. The lowest pulsation amplitude (A = 20%) had a negligible effect on flow resistance, while higher amplitudes amplified pressure drop in the mid-Wo range. When comparing diameters at constant Wo (∼31), smaller pipes (d = 15 mm) exhibited higher spectral energy in both primary and secondary peaks and resulted in greater resistance than larger pipes (d = 25 mm). Spectral analysis using fast Fourier transform (FFT) and continuous wavelet transform (CWT) revealed that Wo = 30–61 corresponds to chaotic flow structures with multiple dominant frequencies, whereas flows at Wo > 61 showed stronger coherence. The findings suggest that selecting specific combinations of pulsation parameters can optimize flow resistance for various industrial needs such as heat transfer enhancement, drag reduction, or internal pipe cleaning.

Heart disease is the leading cause of mortality globally. Electrocardiograms (ECGs) are standard instruments for the examination of heart conditions, but traditional analysis is time-consuming and prone to errors. Novel advances in artificial intelligence have improved ECG classification. However, some limitations remain, such as poor interpretability, computational cost, and class imbalance. This study proposes a novel deep learning algorithm based on Depthwise Separable Residual Attention called DRA-ECG and a customized Adaptive Focal Cross-Entropy (AFCE) loss function for cardiac condition classification. This proposed methodology leverages the Continuous Wavelet Transform (CWT) method to transform 1D raw ECG signals into 2D scalograms to enhance feature representation and training. The proposed customized AFCE loss function incorporated into the DRA-ECG model addresses the class imbalance problem and boost the performance of the model. More so, this study incorporates edge feature detection as a preprocessing technique to denoise and enhance the trainable features of the 2D scalograms for optimal feature representation. The proposed DRA-ECG model achieves a high accuracy of 98.17%, recall of 95.78%, F1-score of 95.82%, and precision of 95.89%. This study shows that the results achieved by the proposed DRA-ECG surpass the current state-of-the-art and existing research works, concerning classification performance and generalization ability in ECG classification, which underlines the effectiveness of the novel AFCE loss function for ensuring high-classification accuracy and robustness. The proposed novel methodology enhances heart disease classification and provides a robust and reliable solution for medical diagnosis, addressing the major drawbacks of existing models.

Accurate classification of sleep stages is crucial for the diagnosis and management of sleep disorders. Conventional approaches for sleep scoring rely on manual annotation or features extracted from EEG signals in the time or frequency domain. This study proposes a novel framework for automated sleep stage scoring using time–frequency analysis based on the wavelet transform. The Sleep-EDF Expanded Database (sleep-cassette recordings) was used for evaluation. The continuous wavelet transform (CWT) generated time–frequency maps that capture both transient and oscillatory patterns across frequency bands relevant to sleep staging. Experimental results demonstrate that the proposed wavelet-based representation, combined with ensemble learning, achieves an overall accuracy of 88.37% and a macro-averaged F1 score of 73.15%, outperforming conventional machine learning methods and exhibiting comparable or superior performance to recent deep learning approaches. These findings highlight the potential of wavelet analysis for robust, interpretable, and clinically applicable sleep stage classification.

Magnetorheological dampers (MRDs) are widely used in the field of engineering vibration control. The accurate identification and prediction of their output forces are crucial for optimizing control strategies. However, traditional analysis methods based on mechanical models and empirical formulas have many limitations. This study proposes an innovative deep-learning approach. First, the continuous wavelet transform (CWT) is employed to convert the one-dimensional signals of MRD output force into two-dimensional time-frequency maps. Then, a convolutional neural network (CNN) is employed for feature extraction and type identification, and a CWT-CNN model is constructed. This model achieves 100% accuracy on the test dataset. In addition, by combining the local feature extraction ability of CNN and the sequence modeling advantage of the long short-term memory network (LSTM), a CNN-LSTM model is built to predict the MRD output force. The results show that compared with CNN and LSTM, the CNN-LSTM model exhibits stronger generalization ability. It outperforms the other models in comprehensive performance, as evaluated by MSE, RMSE, MAE, MAPE, and R2. This study provides an effective technical means for the identification and prediction of MRD output forces and promotes the application and development of deep learning in this field.