Search Results (41)

The generation of synthetic seismic accelerograms is a critical problem in earthquake engineering, where the scarcity of strong-motion records, particularly for high-magnitude and near-fault scenarios, limits the reliability of structural analyses and probabilistic seismic hazard assessments. This paper presents a proof-of-concept wavelet-decomposed conditional Generative Adversarial Network (WD-cGAN) for the synthesis of seismic accelerograms that reproduce the physical and statistical properties of real ground-motion records. Unlike prior GAN-based approaches that rely on Fourier-domain decomposition, the proposed architecture decomposes each training signal into N wavelet sub-bands (experimentally $N=7$, six detail sub-bands D1–D6 and one approximation sub-band A6) using the Daubechies-4 (db4) discrete wavelet transform (DWT), assigning each sub-band to a dedicated discriminator. A novel energy-based weighting scheme ${\alpha }_i$ modulates the relative contribution of each discriminator to the total generator loss, ensuring that physically dominant, low-frequency bands, which carry the bulk of seismic energy, receive proportionally higher training emphasis. Seismic moment magnitude $M_w$ serves as the primary conditioning variable, enabling targeted synthesis for specific hazard scenarios. The model is implemented in Python v3.9 using PyTorch v.2.10 and trained on accelerograms drawn from the Italian INGV/ITACA v4.0 archive. Preliminary evaluation on 500 synthetic accelerograms across five magnitude classes provides evidence that the proposed wavelet-domain multi-discriminator scheme reproduces the essential spectral shape and non-stationary temporal structure of real ground-motion records within the considered magnitude range; full quantitative validation on a larger and more diverse corpus, rigorous comparison with competing methods, and extended multi-parameter conditioning are identified as the principal avenues for future work. Full article

The mathematical characterization of non-stationary signals remains a significant challenge, particularly when impulsive components are obscured by high-dimensional noise and structural coupling. This paper proposes an application-driven mathematical methodology for a learnable discrete wavelet transform (LDWT) that combines classical multi-resolution analysis with task-optimized data-driven adaptivity. Rather than introducing entirely new foundational theory, our approach strategically relaxes constraints from orthogonal wavelet theory within the non-perfect reconstruction filter bank framework, enabling controlled spectral decomposition optimized for supervised fault diagnosis. We introduce a specialized regularization term based on the half-band property to ensure spectral complementarity and minimize cross-band correlation, while a Jacobian-based stabilization approach is formulated to ensure the convergence of filter coefficients during optimization. The proposed algorithmic architecture, LDBRFnet, features a dual-branch encoder system designed to capture the mathematical synergy between sub-band-level global statistics and time-domain local morphology. This dual-view representation effectively mitigates feature leakage and overconfidence in classification. Theoretical analysis and numerical experiments demonstrate that the learned filters satisfy the frequency-shift property and maintain robust spectral partitioning even under low signal-to-noise ratios. Validation on complex vibration datasets confirms that the framework achieves superior diagnostic accuracy (over 95.5%) and computational efficiency, reducing model parameters by 96.7% compared to state-of-the-art baselines. This work provides a generalizable mathematical approach for adaptive signal decomposition and robust pattern recognition in interdisciplinary applications. Full article

Short-term forecasting of carbon dioxide (CO₂) emissions supports near-term mitigation planning, but real-world emission series are nonlinear, non-stationary, and contaminated by multi-scale fluctuations. This paper proposes RVWF-iTransformer, which integrates RIME-optimized variational mode decomposition (RIME-VMD), a trainable one-dimensional wavelet convolution block (WTConv_1d), and a discrete-cosine-transform-based frequency-enhanced channel attention mechanism (FECAM) into an iTransformer backbone. The model was evaluated on daily national CO₂ emissions for China and India using Carbon Monitor and on two public benchmarks (PM_2.5 and ETTH₂) using a chronological 8/2 (train/test) split (with validation taken from the tail of the training segment for early stopping) and horizons $H\in \{5,10,15,20\}$. Predictive results are reported as mean ± standard deviation over 20 independent runs; RVWF-iTransformer yields the lowest errors at longer horizons on China-CO₂ and maintains robust performance under additive Gaussian noise with SNR = 20/10/5 dB. These findings suggest that aligning adaptive decomposition and time–frequency representation learning within a single causal pipeline improves forecasting stability for non-stationary environmental time series. Full article

Myocardial infarction (MI) is a life-threatening cardiovascular disorder caused by a partial or complete interruption of oxygenated blood flow to the myocardium, leading to high mortality rates if not diagnosed promptly. Although electrocardiogram (ECG) signals are widely used due to their non-invasive and low-cost nature, MI-specific abnormalities may be subtle and subject to inter-observer variability. Therefore, reliable artificial intelligence-based decision support systems are essential to enhance diagnostic classification accuracy. In this study, only the Lead II derivation from 12-lead ECG recordings of 52 healthy individuals and 148 MI patients was analyzed. To effectively characterize the non-stationary nature of ECG signals, a hybrid time–frequency feature extraction framework was employed. Five-level intrinsic mode functions and wavelet detail and approximation coefficients were obtained using Empirical Mode Decomposition and Discrete Wavelet Transform with a Daubechies-6 wavelet. From these components, 390 times, nonlinear and complexity-based features were extracted using 23 entropy-driven measures. Particle Swarm Optimization was applied to select the most discriminative feature subset, significantly enhancing classification performance. The optimized features were evaluated using Support Vector Machines, Artificial Neural Networks, k-Nearest Neighbors, and Bagged Tree classifiers. The Bagged Trees classifier achieved the best classification performance with an overall correct classification rate of 97.6%. The results demonstrate that the proposed hybrid feature representation combined with PSO-based selection provides a robust and reliable framework for MI detection, offering strong potential for clinical decision support applications. Full article

Early detection of fire-related risks in lithium-ion batteries (LIBs) remains a critical challenge, as conventional protection mechanisms typically activate only after irreversible degradation or macroscopic failure occurs. In this study, an innovative sensor-based diagnostic framework is proposed for proactive fire prediction in LIBs by simultaneously monitoring low-frequency and high-frequency electrical signatures generated during battery charge–discharge processes. An electromagnetic (EM) antenna sensor and a high-frequency current transformer (HFCT) sensor were employed to capture complementary voltage- and current-based transient signals associated with internal degradation phenomena. Cell-level experiments were conducted under various C-rates and temperature conditions, including high-stress environments, while module-level validation was performed on a 4-series, 1-parallel (4S1P) configuration at a 2C-rate under ambient temperature. Time–frequency characteristics of the measured signals were systematically evaluated using MATLAB-based continuous wavelet transform (CWT) and discrete wavelet transform (DWT) techniques. The results reveal that degradation-induced transient events exhibit non-stationary, impulsive voltage and current signatures with distinct frequency-band localization, which intensify with increasing C-rate, elevated temperature, and aging progression. At the module level, although signal amplitudes were partially attenuated due to current redistribution, characteristic wavelet energy patterns and time–frequency concentrations remained clearly distinguishable, demonstrating the scalability of the proposed approach. The combined EM antenna–HFCT sensing strategy, together with multi-resolution wavelet analysis, enables effective phenomenological differentiation between normal operational noise and incipient internal fault signatures well before conventional thermal or capacity-based indicators become evident. These findings demonstrate feasibility of the proposed method for early-stage fault diagnosis and highlight its potential applicability to advanced battery management systems for proactive fire prevention in large-scale energy storage and electric vehicle applications. Unlike conventional voltage-, temperature-, or gas-based diagnostics, the proposed approach enables the detection of incipient degradation phenomena at the microsecond scale by exploiting complementary low- and high-frequency electrical signatures. This study provides experimental evidence that wavelet-based EM and HFCT sensing can identify MISC-related precursors significantly earlier than conventional battery management indicators. Full article

The early and accurate identification of breast cancer is a significant healthcare issue, largely because the traditional machine learning approaches rely on handcrafted features that are unable to fully capture the spatial and textural complexity found in mammograms. Even with the advancements made possible through deep learning and improvements in diagnostic performance, most computational-aided diagnosis (CAD) systems based on Convolutional Neural Networks (CNNs) still only rely on single-domain features, normally spatial features, while neglecting some important spectral and spatial–spectral features, leading to limitations in generalisability, redundancy, and loss of performative interpretability. Inspired by these limitations, this research proposes MERGE, a novel CAD framework that combines spatial, spectral, and spatial–spectral information—all part of a single multistage architecture taking advantage of three fine-tuned CNN models (ResNet-50, Xception, and Inception). This system utilises Discrete Stationary Wavelet Transform (DSWT) to enhance spectral–spatial features; Discrete Cosine Transform (DCT) to fuse the features optimally, resulting in enhanced spatial and spatial–spectral representations; and, finally, Non-Negative Matrix Factorisation (NNMF) for reduced-dimensional features. Finally, the Linear Discriminant Analysis (LDA), support vector machine (SVM), and k-nearest neighbours (KNN) classifiers provide a robust diagnosis. Using the INBreast and MIAS datasets in evaluations of the experimental research design, evaluation metrics of accuracy, sensitivity, specificity, and AUC were around 99%, with performance surpassing state-of-the-art paradigms. The findings of the suggested MERGE indicate significant promise as a dependable and effective diagnostic tool, enhancing the consistency and interpretability of breast cancer screening results. Full article

Gravitational-wave (GW) detection has become a significant area of research following the first successful observation by the Laser Interferometer Gravitational-Wave Observatory (LIGO). The detection of signals emerging from binary black hole (BBH) mergers have challenges due to the presence of non-Gaussian and non-stationary noise in observational data. Using traditional matched filtering techniques to detect BBH merging are computationally expensive and may not generalize well to unexpected GW events. As a result, deep learning-based methods have emerged as powerful alternatives for robust GW signal detection. In this study, we propose a novel Transformer-based architecture that introduces energy-aware modulation into the attention mechanism through dual-energy attention masks. In the proposed framework, Q-transform and discrete wavelet transform (DWT) are employed to extract time–frequency energy representations from gravitational-wave signals which are fused into energy masks that dynamically guide the Transformer encoder. In parallel, the raw one-dimensional signal is used directly as input and segmented into temporal patches, which enables the model to leverage both learned representations and physically grounded priors. This proposed architecture allows the model to focus on energy-rich and informative regions of the signal in order to enhance the robustness of the model under realistic noise conditions. Experimental results on BBH datasets embedded in real LIGO noise show that EDAT-BBH outperforms CNN-based and standard Transformer-based approaches, achieving an accuracy of 0.9953, a recall of 0.9950, an F1-score of 0.9953, and an AUC of 0.9999. These findings demonstrate the effectiveness of energy-modulated attention in improving both the interpretability and performance of deep learning models for gravitational-wave signal classification. Full article

The article presents Modified Scaled Adaptive Daqrouq Wavelet (MSADW) as an autonomous wavelet framework to overcome the analysis obstacles of traditional wavelets (Morlet and Daubechies) for signals with non-stationary characteristics. MSADW adjusts its waveform shape and frequency in real time based on the specific characteristics of the signal, allowing it to outperform conventional wavelet methods. The system reaches adaptability through three core methods featuring gradient-dependent scale adjustments for fast transient detection and smooth regions, and instantaneous frequency monitoring achieved by a combination of STFT and Hilbert transforms and an iterative error reduction process using gradient descent and genetic algorithms. Continuous Wavelet Transform (CWT) combined with Discrete Wavelet Transform (DWT) extracts features from ECG and speech signals. Throughout this process, MSADW maintains great time precision to detect transients as well as maintain sensitivity for the audio’s base stability. Testing MSADW in practical use reveals its superior performance because it detects R-peaks accurately within 0.01 s through zero-crossing methods, which combine P/T-wave detection with effective ECG signal segmentation and noise-free reconstructed speech ($\mathit{MSE:} 1.17\times {10}^{-31}$). The localized parameterization framework of MSADW, enabled by feedback refinement, fulfills missing aspects in biomedical signal evaluation and creates space for low-cost real-time evaluation methods for medical devices and arrhythmia and ischemic detection platforms. The theoretical backbone for MSADW establishes itself because this work shows how wavelet analysis can transition toward managing non-stationary and noise-prone domains. Full article

As a form of non-destructive testing, eddy current testing is widely used for detecting surface micro-damage on metal components in sectors such as aerospace. Conventional frequency-domain analysis techniques often fail to effectively extract defect-related features from non-stationary eddy current signals. This paper proposes an ECT system based on the Discrete Wavelet Transform to address this limitation. In hardware design, the system employs a DDS to generate a 1 MHz excitation signal for the probe. High-precision acquisition of defect response signals is achieved using an IQ demodulator and a 24-bit ADC. For signal processing, the Haar wavelet is applied for single-level decomposition. This method successfully isolates the defect response signal within the high-frequency detail coefficients. Experimental results demonstrate that for a metal surface notch with a depth of 1 mm, the system significantly improves the SNR, resulting in a ΔSNR improvement of 3.64 dB, which is 0.36 dB higher than that achieved using time-domain processing. Full article

Landslide displacement prediction is crucial for disaster mitigation, yet traditional methods often fail to capture the complex, non-stationary spatiotemporal dynamics of slope evolution. This study introduces an enhanced prediction framework that integrates multi-scale signal processing with dynamic, geology-aware graph modeling. The proposed methodology first employs the Maximum Overlap Discrete Wavelet Transform (MODWT) to denoise raw Global Navigation Satellite System (GNSS)-monitored displacement time series data, enhancing the underlying deformation features. Subsequently, a geology-aware graph is constructed, using the temporal correlation of displacement series as a practical proxy for physical relatedness between monitoring nodes. The framework’s core innovation lies in a dynamic graph optimization model with low-rank constraints, which adaptively refines the graph topology to reflect time-varying inter-sensor dependencies driven by factors like mining activities. Experiments conducted on a real-world dataset from an active open-pit mine demonstrate the framework’s superior performance. The DCRNN-proposed model achieved the highest accuracy among eight competing models, recording a Root Mean Square Error (RMSE) of 2.773 mm in the Vertical direction, a 39.1% reduction compared to its baseline. This study validates that the proposed dynamic graph optimization approach provides a robust and significantly more accurate solution for landslide prediction in complex, real-world engineering environments. Full article

Inter-turn short-circuit (ITSC) faults in permanent magnet synchronous machines (PMSMs) present a significant reliability challenge in electric vehicle (EV) drivetrains, particularly under non-stationary operating conditions characterized by inverter-driven transients, variable loads, and magnetic saturation. Existing diagnostic approaches, including motor current signature analysis (MCSA) and wavelet-based methods, are primarily designed for steady-state conditions and rely on manual feature selection, limiting their applicability in real-time embedded systems. Furthermore, the lack of publicly available, high-fidelity datasets capturing the transient dynamics and nonlinear flux-linkage behaviors of PMSMs under fault conditions poses an additional barrier to developing data-driven diagnostic solutions. To address these challenges, this study introduces a simulation framework that generates a comprehensive dataset using finite element method (FEM) models, incorporating magnetic saturation effects and inverter-driven transients across diverse EV operating scenarios. Time-frequency features extracted via Discrete Wavelet Transform (DWT) from stator current signals are used to train a Transformer model for automated ITSC fault detection. The Transformer model, leveraging self-attention mechanisms, captures both local transient patterns and long-range dependencies within the time-frequency feature space. This architecture operates without sequential processing, in contrast to recurrent models such as LSTM or RNN models, enabling efficient inference with a relatively low parameter count, which is advantageous for embedded applications. The proposed model achieves 97% validation accuracy on simulated data, demonstrating its potential for real-time PMSM fault detection. Additionally, the provided dataset and methodology contribute to the facilitation of reproducible research in ITSC diagnostics under realistic EV operating conditions. Full article

This study investigates the occurrence and distribution of geomagnetic pulsations (Pc2–Pc5) over South America during 2014, analyzing their dependence on magnetic latitude, local time, and geomagnetic activity. Geomagnetic field data were obtained from the Embrace magnetometer network, which spans Brazil and Argentina and includes regions influenced by the Equatorial Electrojet (EEJ) and the South Atlantic Magnetic Anomaly (SAMA). Both continuous and discrete wavelet transforms (CWT and DWT) were employed to analyze non-stationary signals and reconstruct pulsation activity during quiet and disturbed geomagnetic periods. The results reveal that Pc5 and Pc3 pulsations exhibit a pronounced diurnal peak around local noon, with significantly stronger and more widespread activity under storm conditions. Spatial analyses highlight localized enhancements near the dip equator during quiet times and broader latitudinal spread during geomagnetic disturbances. These findings underscore the strong modulation of pulsation activity by geomagnetic conditions and offer new insights into wave behavior at low and mid-latitudes. This work contributes to understanding magnetosphere–ionosphere coupling and has implications for space weather prediction and geomagnetically induced current (GIC) risk assessment in the South American sector. Full article

(1) Background: Due to its imaging principle, OCT generates images laden with significant speckle noise. The quality of OCT images is a crucial factor influencing diagnostic effectiveness, highlighting the importance of OCT image denoising. (2) Methods: The OCT image undergoes a Discrete Wavelet Transform (DWT) to decompose it into multiple scales, isolating high-frequency wavelet coefficients that encapsulate fine texture details. These high-frequency coefficients are further processed using a Shift-Invariant Wavelet Transform (SWT) to generate an additional set of coefficients, ensuring an enhanced feature preservation and reduced artifacts. Both the original DWT high-frequency coefficients and their SWT-transformed counterparts are independently denoised using a Deep Neural Convolutional Network (DnCNN). This dual-pathway approach leverages the complementary strengths of both transform domains to suppress noise effectively. The denoised outputs from the two pathways are fused using a correlation-based strategy. This step ensures the optimal integration of texture features by weighting the contributions of each pathway according to their correlation with the original image, preserving critical diagnostic information. Finally, the Inverse Wavelet Transform is applied to the fused coefficients to reconstruct the denoised OCT image in the spatial domain. This reconstruction step maintains structural integrity and enhances diagnostic clarity by preserving essential spatial features. (3) Results: The MSE, PSNR, and SSIM indices of the proposed algorithm in this paper were 4.9052, 44.8603, and 0.9514, respectively, achieving commendable results compared to other algorithms. The Sobel, Prewitt, and Canny operators were utilized for edge detection on images, which validated the enhancement effect of the proposed algorithm on image edges. (4) Conclusions: The proposed algorithm in this paper exhibits an exceptional performance in noise suppression and detail preservation, demonstrating its potential application in OCT image denoising. Future research can further explore the adaptability and optimization directions of this algorithm in complex noise environments, aiming to provide more theoretical support and practical evidence for enhancing OCT image quality. Full article

Background and Objectives: Lie detection is crucial in domains such as security, law enforcement, and clinical assessments. Traditional methods suffer from reliability issues and susceptibility to countermeasures. In recent years, electroencephalography (EEG) and particularly the Event-Related Potential (ERP) P300 component have gained prominence for identifying concealed information. This systematic review aims to evaluate recent studies (2017–2024) on EEG-based lie detection using ERP P300 responses, especially in relation to recognized and unrecognized face stimuli. The goal is to summarize commonly used EEG signal processing techniques, feature extraction methods, and classification algorithms, identifying those that yield the highest accuracy in lie detection tasks. Methods: This review followed PRISMA guidelines for systematic reviews. A comprehensive literature search was conducted using IEEE Xplore, Web of Science, Scopus, and Google Scholar, restricted to English-language articles from 2017 to 2024. Studies were included if they focused on EEG-based lie detection, utilized experimental protocols like Concealed Information Test (CIT), Guilty Knowledge Test (GKT), or Deceit Identification Test (DIT), and evaluated classification accuracy using ERP P300 components. Results: CIT with ERP P300 was the most frequently employed protocol. The most used preprocessing method was Bandpass Filtering (BPF), and the Discrete Wavelet Transform (DWT) emerged as the preferred feature extraction technique due to its suitability for non-stationary EEG signals. Among classification algorithms, Support Vector Machine (SVM), Linear Discriminant Analysis (LDA), and Convolutional Neural Networks (CNN) were frequently utilized. These findings demonstrate the effectiveness of hybrid and deep learning-based models in enhancing classification performance. Conclusions: EEG-based lie detection, particularly using the ERP P300 response to face recognition tasks, shows promising accuracy and robustness compared to traditional polygraph methods. Combining advanced signal processing methods with machine learning and deep learning classifiers significantly improves performance. This review identifies the most effective methodologies and suggests that future research should focus on real-time applications, cross-individual generalization, and reducing system complexity to facilitate broader adoption. Full article

This work presents a vibration-based non-contact monitoring method to classify the physical activity of a mouse inside a home cage. A novel tuned-beam sensing device is developed to measure low-amplitude activity-induced cage vibrations. The sensing device uses a mechanical beam structure to enhance a six-axis IMU that increases the signal-to-noise ratio (SNR) by 20 to 40 times in a relevant environment. A sophisticated classification algorithm is developed to process vibration sequences with a variable time frame that utilizes multi-level discrete wavelet transformation (MLDWT) to extract time–frequency features and optimize signal properties. The extracted features are classified by a convolutional neural network–long short-term memory (CNN-LSTM) machine learning model to determine the activity class. The ground truth is obtained with a camera-based system using EthoVision XT from Noldus and a custom post-processor. The method is developed on a dataset containing 300 h of vibration measurements with camera-based reference and includes two separate home cages and two individual mice. The method classifies the activity types Resting, Stationary Activity, Walking, Activity in Feeder, and Drinking with an accuracy of 86.81% and an average F1 score of 0.798 using a 9 s time frame. In long-term monitoring, the proposed method reproduces behavioral patterns such as sleep and acclimatization as accurately as the reference method, enabling home cage monitoring in the husbandry environment with a low-cost sensor. Full article