Signal Preprocessing Research Articles

Enhancing Real-Time Cursor Control with Motor Imagery and Deep Neural Networks for Brain–Computer Interfaces

This paper advances real-time cursor control for individuals with motor impairments through a novel brain–computer interface (BCI) system based solely on motor imagery. We introduce an enhanced deep neural network (DNN) classifier integrated with a Four-Class Iterative Filtering (FCIF) technique for efficient preprocessing of neural signals. The underlying approach is the Four-Class Filter Bank Common Spatial Pattern (FCFBCSP) and it utilizes a customized filter bank for robust feature extraction, thereby significantly improving signal quality and cursor control responsiveness. Extensive testing under varied conditions demonstrates that our system achieves an average classification accuracy of 89.1% and response times of 663 milliseconds, illustrating high precision in feature discrimination. Evaluations using metrics such as Recall, Precision, and F1-Score confirm the system’s effectiveness and accuracy in practical applications, making it a valuable tool for enhancing accessibility for individuals with motor disabilities.

Evaluating the Impact of Windowing Techniques on Fourier Transform-Preprocessed Signals for Deep Learning-Based ECG Classification

(1) Background: Arrhythmias, or irregular heart rhythms, are a prevalent cardiovascular condition and are diagnosed using electrocardiogram (ECG) signals. Advances in deep learning have enabled automated analysis of these signals. However, the effectiveness of deep learning models depends greatly on the quality of signal preprocessing. This study evaluated the impact of different windowing techniques applied to Fourier transform-preprocessed ECG signals on the classification accuracy of deep learning models. (2) Methods: We applied three windowing techniques—Hamming, Hann, and Blackman—to transform ECG signals into the frequency domain. A one-dimensional convolutional neural network was employed to classify the ECG signals into five arrhythmia categories based on features extracted from each windowed signal. (3) Results: The Blackman window yielded the highest classification accuracy, with improved signal-to-noise ratio and reduced spectral leakage compared to the Hamming and Hann windows. (4) Conclusions: The choice of windowing technique significantly influences the effectiveness of deep learning models in ECG classification. Future studies should explore additional preprocessing methods and their clinical applications.

Wear fault diagnosis in hydro-turbine via the incorporation of the IWSO algorithm optimized CNN-LSTM neural network

When a hydropower unit operates in a sediment-laden river, the sediment accelerates hydro-turbine wear, leading to efficiency loss or even shutdown. Therefore, wear fault diagnosis is crucial for its safe and stable operation. A hydro-turbine wear fault diagnosis method based on improved WT (wavelet threshold algorithm) preprocessing combined with IWSO (improved white shark optimizer) optimized CNN-LSTM (convolutional neural network-long-short term memory) is proposed. The improved WT algorithm is utilized to denoise the preprocessing of the original signals. Chaotic mapping, bird flock search, and cosine elite variation strategies are introduced to enhance the WSO algorithm's robust performance, and the CNN-LSTM model's hyperparameters are optimized using the IWSO algorithm to improve the diagnostic performance. The experimental results show that the accuracy of the proposed method reaches 96.2%, which is 8.9% higher than that of the IWSO-CNN-LSTM model without denoising. The study also found that the diagnostic accuracy of hydro-turbine wear faults increased with increasing sediment concentration in the water. This study can supplement the existing hydro-turbine condition monitoring and fault diagnosis system. Meanwhile, diagnosing wear faults in hydro-turbines can improve power generation efficiency and quality and minimize resource consumption.

Changes in P300 amplitude to negative emotional stimuli correlate with treatment responsiveness to sertraline in adolescents with depression

ObjectiveAdolescents with depression is characterized by high rates of recurrence and functional impairment, with a significant association with suicide risk. Antidepressants are commonly prescribed to treat depression, yet few reproducible neurobiological markers for depression and antidepressant treatment response have been identified. Therefore, discovering a stable and reliable neurobiological marker holds significant value for both the clinical diagnosis and treatment of depression in adolescents. MethodsOne hundred and seven patients with major depressive disorder (MDD group, 30 males, 77 females, mean age: 14.80 years), and 25 healthy subjects (HC group, 13 males, 12 females, mean age: 15.72 years) were recruited to perform a two-choice oddball task related to negative emotional cues. All participants completed a self-administered questionnaire to gather demographic information. A trained psychiatrist administered the Hamilton Depression Scale (HAMD-17) to assess depression severity. Of the 107 adolescents with depression, 61 received antidepressant medication for 8 weeks, and 61 of these patients were followed up. Multichannel EEG was recorded continuously from 64 scalp electrodes using the Curry 8 system. EEG signal preprocessing and analysis was performed offline using the EEGLAB toolbox in MATLAB. The ERP component characteristics associated with emotional processing were extracted from the difference waves and statistically analyzed. ResultsAdolescents with depression exhibited significantly larger P300 amplitudes than healthy controls in response to both neutral and negative emotional cues. Following sertraline treatment, both depression scores and P300 amplitudes decreased significantly in adolescents with depression. Moreover, a strong positive correlation was observed between changes in depression scores and changes in P300 amplitude in response to negative emotional cues before and after treatment. ConclusionsChanges in neural reactivity to negative emotional stimuli among adolescents with depression can be selectively modulated by sertraline and are significantly associated with improvements in depressive symptoms. SignificanceChanges in P300 amplitude to negative emotional stimuli significantly correlate with treatment responsiveness to sertraline in adolescents with depression.

Diagnosis of Schizophrenia Using EEG Sensor Data: A Novel Approach with Automated Log Energy-Based Empirical Wavelet Reconstruction and Cepstral Features.

Schizophrenia (SZ) is a severe mental disorder characterised by disruptions in cognition, behaviour, and perception, significantly impacting an individual's life. Traditional SZ diagnosis methods are labour-intensive and prone to errors. This study presents an innovative automated approach for detecting SZ acquired through electroencephalogram (EEG) sensor signals, aiming to improve diagnostic efficiency and accuracy. We utilised Fast Independent Component Analysis to remove artefacts from raw EEG sensor data. A novel Automated Log Energy-based Empirical Wavelet Reconstruction (ALEEWR) technique was introduced to reconstruct decomposed modes based on their variability, ensuring effective extraction of meaningful EEG signatures. Cepstral-based features-cepstral activity, cepstral mobility, and cepstral complexity-were used to capture the power, rate of change, and irregularity of the cepstrum of preprocessed EEG signals. ANOVA-based feature selection was applied to refine these features before classification using the K-Nearest Neighbour (KNN) algorithm. Our approach achieved an exceptional accuracy of 99.4%, significantly surpassing previous methods. The proposed ALEEWR and cepstral analysis demonstrated high precision, sensitivity, and specificity in the automated diagnosis of schizophrenia. This study introduces a highly accurate and efficient method for SZ detection using EEG technology. The proposed techniques offer significant improvements in diagnostic accuracy, with potential implications for enhancing SZ diagnosis and patient care through automated systems.

Wind turbine condition monitoring dataset of Fraunhofer LBF

Fraunhofer wind turbine dataset contains monitoring data from a 750 W wind turbine, including accelerometers and tachometer, to capture structural response, bearing vibrations and rotational velocity. Additionally, temperatures, wind speed and wind direction have been measured, while weather conditions have been acquired from selected sources. Various damage scenarios, including mass imbalance, and aerodynamic imbalance as well as damages on bearings’ outer race, inner race and roller element have been implemented. The availability of time series data makes the dataset well suited for both machine learning and signal processing-based condition monitoring applications. The availability of heterogeneous sensors has created a dataset particularly suited for information fusion, data fusion, multi-sensor approaches, and holistic monitoring. Experiments were conducted in real-world conditions outside of a controlled laboratory environment, thereby introducing challenges such as variable rotor speed, noise, overloads, and other environmental factors. Consequently, the dataset is qualified for tasks involving uncertainty quantification and signal pre-processing. This document will detail the test equipment, experimental procedures, simulated damage cases and measurement parameters.

Physics-based digital twin updating and twin-based explainable crack identification of mechanical lap joint

The mechanical joints, including the lap joint, weld, bolt, and pin, are vulnerable to fatigue failure because of stress concentration and internal flaws. Digital twin (DTw) strategies were proposed to prevent catastrophic system failure by fatigue damage in mechanical joints. In previous studies, the data-driven approach, such as deep learning and machine learning were utilized to estimate severity of the damage. However, it needs to improve its prediction accuracy because of insufficient data and physical interpretability. In this study, the physics-based digital twin model updating and twin-based crack identification of fatigue damage in riveted lap joints were proposed using lamb waves with consideration of uncertain crack growth path. The proposed approach is based on three techniques; (i) Data pre-processing, including filtering and optimization-based signal synchronization, (ii) Lamb-wave propagation analysis with sensor dynamics model and uncertain crack path, and (iii) Optimization based physics-based model updating and inference. In data pre-processing, the excitation frequency magnitude and truncation time are estimated using the observed actuator signal in the Lamb-wave test. The sensor dynamic model and model parameters are updated using the Bayesian optimization method to minimize both the errors in the predicted (y^t) and observed (yt) wave signal and the errors in the inferred (l*) and observed (l) crack length. The crack growth path is sampled based on angular and spline schemes to consider uncertain crack propagation paths. The validity of the proposed method is demonstrated using an open data set (2019 PHM society data challenge). The results conclude that the proposed digital twin approach can improve estimation accuracy considering both the crack growth path and sensor dynamics model.

Enhanced diagnostics for generalized anxiety disorder: leveraging differential channel and functional connectivity features based on frontal EEG signals

Generalized Anxiety Disorder (GAD) is a chronic anxiety condition characterized by persistent excessive worry, anxiety, and fear. Current diagnostic practices primarily rely on clinicians’ subjective assessments and experience, highlighting a need for more objective and reliable methods. This study collected 10-minute resting-state electroencephalogram (EEG) from 45 GAD patients and 36 healthy controls (HC), focusing on six frontal EEG channels for preprocessing, data segmentation, and frequency band division. Innovatively, this study introduced the “Differential Channel” method, which enhances classification performance by enhancing the information related to anxiety from the data, thereby highlighting signal differences. Utilizing the preprocessed EEG signals, undirected functional connectivity features (Phase Lag Index, Pearson Correlation Coefficient, and Mutual Information) and directed functional connectivity features (Partial Directed Coherence) were extracted. Multiple machine learning models were applied to distinguish between GAD patients and HC. The results show that the Deep Forest classifier achieves excellent performance with a 12-second time window of DiffFeature. In particular, the classification of GAD and HC was successfully obtained by combining OriFeature and DiffFeature on Mutual Information with a maximum accuracy of 98.08%. Furthermore, it was observed that undirected functional connectivity features significantly outperformed directed functional connectivity when fewer frontal channels were used. Overall, the methodologies developed in this study offer accurate and practical identification strategies for the early screening and clinical diagnosis of GAD, offering the necessary theoretical and technical support for further enhancing the portability of EEG devices.

Emotion recognition based on a limited number of multimodal physiological signals channels

Emotion constitutes a higher-order cognitive process, moreover, physiological signals can offer a more objective and realistically reflection of human emotional state. Electroencephalograph (EEG) signals, controlled by the central nervous system, are highly sensitive to emotional state fluctuations. Current research primarily focuses on recognizing emotions through feature extraction from multi-channel EEG signals. However, the acquisition of multi-channel EEG signals is hindered by difficulties acquiring from hair occlusion. The objective of this study is to construct a real-time physiological-emotion recognition model based on 2-channel EEG signals from the forehead without hair occlusion combined with 3-channel peripheral physiological (PPS) signals, employing signal preprocessing, feature extraction, feature normalization, and the ensemble learning algorithm Light Gradient Boosting Machine (LightGBM). In this paper, thirty participants sequentially viewed fourteen virtual reality (VR) scenes, rated for valence and arousal using the Self-Assessment Manikin (SAM) scale after each scene. The 2-channel EEG signals and 3-channel PPS signals were recorded synchronously from forehead by VR Head-Mounted Displays (HMD) Bio Pad developed in our research center. Our proposed model achieved median accuracy of 89.68% for valence, 90.11% for arousal in binary classification, and 84.93% for four classification of emotion recognition based on the collected samples. Furthermore, verification result of the proposed method on a public database (DEAP) achieved accuracy of 84.03%, 84.37%, and 72.23% on valence, arousal, and four classification, respectively. The proposed physiological-emotion recognition model, utilizing a limited number of multimodal physiological signals channels from wearable devices, demonstrates higher accuracy compared with the results in some literature, suggesting its potential for real-time implementation due to limited signal channels and low runtime.

Characterization of internal defects in cylindrical components using laser ultrasonic method with a modified SAFT algorithm

The health monitoring of cylindrical elements is particularly important for the safe operation of industrial production, because cylinders bear a lot of support and transmission work. Normal non-destructive evaluation techniques need special designs to suit high curvature surfaces of cylinders. Laser ultrasonic (LU) method can provide a remote and non-destructive inspection solution to solid cylinders due to its flexible adaptability to complex structures and strong penetration depth. However, constrained by the common problems of optical detection systems, the detected ultrasonic signals will suffer low signal-to-noise ratio and bad resolution from poor sample surface quality. Thus, a modified synthetic aperture focusing technique (SAFT) optimized for cylindrical components is proposed to improve the detectability of LU on small and buried defects. Mode conversion wave signals obtained by multiple separate excitations are used in SAFT imaging for the reconstruction of the defects under surface profile correction. To reduce the influence of incident waves such as direct surface acoustic wave, besides common difference method, adjacent wave subtraction algorithm based on cross-correlation is used for signal preprocessing, suppressing the incident waves and exaggerating the mode conversion waves. In numerical simulation, internal defects with diameters from 0.6 to 0.2 mm with various buried depths are visualized and accurately located via the optimized SAFT algorithm using mode conversion waves. For validation, a circumferential scanning system is established in LU experiment and internal defects from 0.8 to 0.4 mm in diameter inside solid cylinders are successfully detected with precise location. The results elucidate the reliability of characterizing the location of small internal buried defects in solid cylinder structures through LU-SAFT imaging.

Piercing the Shadows: Exploring the Influence of Signal Preprocessing on Interpreting Ultrasensitive Bioelectronic Sensor Data.

The development of ultrasensitive electronic sensors for in vitro diagnostics is essential for the reliable monitoring of asymptomatic individuals before illness proliferation or progression. These platforms are increasingly valued for their potential to enable timely diagnosis and swift prognosis of infectious or progressive diseases. Typically, the responses from these analytical tools are recorded as digital signals, with electronic data offering simpler processing compared to spectral and optical data. However, preprocessing electronic data from potentiometric biosensor arrays is still in its infancy compared to more established optical technologies. This study utilized the Single-Molecule with a Large Transistor (SiMoT) array, which has achieved a Technology Readiness Level of 5, to explore the impact of data preprocessing on electronic biosensor outcomes. A dataset consisting of plasma and cyst fluid samples from 37 patients with pancreatic precursor cyst lesions was analyzed. The findings revealed that standard signal preprocessing can produce misleading conclusions due to artifacts introduced by mathematical transformations. The study offers strategies to mitigate these effects, ensuring that data interpretation remains accurate and reflective of the underlying biochemical information in the samples.

Convolutional neural network framework for EEG-based ADHD diagnosis in children.

Attention-deficit hyperactivity disorder (ADHD) stands as a significant psychiatric and neuro-developmental disorder with global prevalence. The prevalence of ADHD among school children in India is estimated to range from 5% to 8%. However, certain studies have reported higher prevalence rates, reaching as high as 11%. Utilizing electroencephalography (EEG) signals for the early detection and classification of ADHD in children is crucial. In this study, we introduce a CNN architecture characterized by its simplicity, comprising solely two convolutional layers. Our approach involves pre-processing EEG signals through a band-pass filter and segmenting them into 5-s frames. Following this, the frames undergo normalization and canonical correlation analysis. Subsequently, the proposed CNN architecture is employed for training and testing purposes. Our methodology yields remarkable results, with 100% accuracy, sensitivity, and specificity when utilizing the complete 19-channel EEG signals for diagnosing ADHD in children. However, employing the entire set of EEG channels presents challenges related to the computational complexity. Therefore, we investigate the feasibility of using only frontal brain EEG channels for ADHD detection, which yields an accuracy of 99.08%. The proposed method yields high accuracy and is easy to implement, hence, it has the potential for widespread practical deployment to diagnose ADHD.

Online chatter detection via lightweight deep learning framework with efficient signal pre-processing

Deep learning methods have attracted considerable attention in the field of chatter detection due to their exceptional feature extraction and identification capabilities. However, it is challenging to adapt these methods to diverse machining conditions with limited sensor resources. In addition, the computational efficiency of these methods is compromised by their complexity, resulting in a reduced real-time performance for online chatter detection. To address them, this paper proposes a novel deep learning-based framework that combines an efficient signal pre-processing method with a novel lightweight parallel convolutional neural network (LPNN). A fast continuous wavelet transform algorithm is applied to the raw signal, effectively generating time–frequency maps that represent chatter characteristics. The LPNN incorporates the condensed block as a feature extraction unit, which is constructed using depthwise separable convolution. The introduction of the condensed block and the parallel structure increases the depth and width of the designed model while maintaining fewer parameters and higher computational efficiency. A transfer learning strategy is adopted to train the model, reducing the training time and data requirements for online chatter detection. Experimental results using large open-source turning datasets demonstrate the effectiveness of the proposed method in identifying chatter under various machining conditions. The proposed method outperforms other methods with a classification accuracy of 96.5% on the test set. Furthermore, the performance of the proposed method for online chatter detection is validated in our milling experiments, exhibiting enhanced real-time performance compared to other methods.

Non-Intrusive System for Honeybee Recognition Based on Audio Signals and Maximum Likelihood Classification by Autoencoder.

Artificial intelligence and Internet of Things are playing an increasingly important role in monitoring beehives. In this paper, we propose a method for automatic recognition of honeybee type by analyzing the sound generated by worker bees and drone bees during their flight close to an entrance to a beehive. We conducted a wide comparative study to determine the most effective preprocessing of audio signals for the detection problem. We compared the results for several different methods for signal representation in the frequency domain, including mel-frequency cepstral coefficients (MFCCs), gammatone cepstral coefficients (GTCCs), the multiple signal classification method (MUSIC) and parametric estimation of power spectral density (PSD) by the Burg algorithm. The coefficients serve as inputs for an autoencoder neural network to discriminate drone bees from worker bees. The classification is based on the reconstruction error of the signal representations produced by the autoencoder. We propose a novel approach to class separation by the autoencoder neural network with various thresholds between decision areas, including the maximum likelihood threshold for the reconstruction error. By classifying real-life signals, we demonstrated that it is possible to differentiate drone bees and worker bees based solely on audio signals. The attained level of detection accuracy enables the creation of an efficient automatic system for beekeepers.

HyEpiSeiD: a hybrid convolutional neural network and gated recurrent unit model for epileptic seizure detection from electroencephalogram signals

Epileptic seizure (ES) detection is an active research area, that aims at patient-specific ES detection with high accuracy from electroencephalogram (EEG) signals. The early detection of seizure is crucial for timely medical intervention and prevention of further injuries of the patients. This work proposes a robust deep learning framework called HyEpiSeiD that extracts self-trained features from the pre-processed EEG signals using a hybrid combination of convolutional neural network followed by two gated recurrent unit layers and performs prediction based on those extracted features. The proposed HyEpiSeiD framework is evaluated on two public datasets, the UCI Epilepsy and Mendeley datasets. The proposed HyEpiSeiD model achieved 99.01% and 97.50% classification accuracy, respectively, outperforming most of the state-of-the-art methods in epilepsy detection domain.

Exploring Computational Techniques in Preprocessing Neonatal Physiological Signals for Detecting Adverse Outcomes: Scoping Review.

Computational signal preprocessing is a prerequisite for developing data-driven predictive models for clinical decision support. Thus, identifying the best practices that adhere to clinical principles is critical to ensure transparency and reproducibility to drive clinical adoption. It further fosters reproducible, ethical, and reliable conduct of studies. This procedure is also crucial for setting up a software quality management system to ensure regulatory compliance in developing software as a medical device aimed at early preclinical detection of clinical deterioration. This scoping review focuses on the neonatal intensive care unit setting and summarizes the state-of-the-art computational methods used for preprocessing neonatal clinical physiological signals; these signals are used for the development of machine learning models to predict the risk of adverse outcomes. Five databases (PubMed, Web of Science, Scopus, IEEE, and ACM Digital Library) were searched using a combination of keywords and MeSH (Medical Subject Headings) terms. A total of 3585 papers from 2013 to January 2023 were identified based on the defined search terms and inclusion criteria. After removing duplicates, 2994 (83.51%) papers were screened by title and abstract, and 81 (0.03%) were selected for full-text review. Of these, 52 (64%) were eligible for inclusion in the detailed analysis. Of the 52 articles reviewed, 24 (46%) studies focused on diagnostic models, while the remainder (n=28, 54%) focused on prognostic models. The analysis conducted in these studies involved various physiological signals, with electrocardiograms being the most prevalent. Different programming languages were used, with MATLAB and Python being notable. The monitoring and capturing of physiological data used diverse systems, impacting data quality and introducing study heterogeneity. Outcomes of interest included sepsis, apnea, bradycardia, mortality, necrotizing enterocolitis, and hypoxic-ischemic encephalopathy, with some studies analyzing combinations of adverse outcomes. We found a partial or complete lack of transparency in reporting the setting and the methods used for signal preprocessing. This includes reporting methods to handle missing data, segment size for considered analysis, and details regarding the modification of the state-of-the-art methods for physiological signal processing to align with the clinical principles for neonates. Only 7 (13%) of the 52 reviewed studies reported all the recommended preprocessing steps, which could have impacts on the downstream analysis. The review found heterogeneity in the techniques used and inconsistent reporting of parameters and procedures used for preprocessing neonatal physiological signals, which is necessary to confirm adherence to clinical and software quality management system practices, usefulness, and choice of best practices. Enhancing transparency in reporting and standardizing procedures will boost study interpretation and reproducibility and expedite clinical adoption, instilling confidence in the research findings and streamlining the translation of research outcomes into clinical practice, ultimately contributing to the advancement of neonatal care and patient outcomes.

Human hand gesture recognition using fast Fourier transform with coot optimization based on deep neural network

ABSTRACT Hand motion detection is particularly important for managing the movement of individuals who have limbs amputated. The existing algorithm is complex, time-consuming and difficult to achieve better accuracy. A DNN is suggested to recognize human hand movements in order to get over these problems.Initially, the raw input EMG signal is captured then the signal is pre-processed using high-pass Butterworth filter and low-pass filter which is utilized to eliminate the noise present in the signal. After that pre-processed EMG signal is segmented using sliding window which is used for solving the issue of overlapping. Then the features are extracted from the segmented signal using Fast Fourier Transform. Then selected the appropriate and optimal number of features from the feature subset using coot optimization algorithm. After that selected features are given as input for deep neural network classifier for recognizing the hand movements of human. The simulation analysis shows that the proposed method obtain 95% accuracy, 0.05% error, precision is 94%, and specificity is 92%.The simulation analysis shows that the developed approach attain better performance compared to other existing approaches. This prediction model helps in controlling the movement of amputee patients suffering from disable hand motion and improve their living standard.

Acoustic modulation signal recognition based on endpoint detection

The modulation recognition technology for acoustic signals holds significant research importance in signal demodulation and communication signal reconnaissance, serving as a crucial component and key aspect. This paper investigates the modulation recognition technology for acoustic signals (< 20 kHz) from the perspectives of signal preprocessing and feature extraction. Firstly, it selects seven modulation signals 2ASK, 4ASK, 2FSK, 4FSK, 2PSK, 4PSK, and OFDM as recognition targets and systematically compares the effectiveness of four different endpoint detection algorithms in modulation signal recognition. To further enhance the performance of the short-time energy entropy ratio algorithm, this study introduces three different noise reduction algorithms for optimization. Finally, to accurately identify and distinguish between 2 and 4FSK signals, this study optimizes the related algorithms of the cyclic spectrum by using the kurtosis coefficient value Kur of the cyclic spectrum parameter matrix when the cyclic frequency α = 0 to differentiate between these two signals. The results show that at SNR of 4 dB, the proposed modulation recognition algorithm can effectively distinguish between these two signals, achieving a recognition accuracy of over 99%.

Adaptive spectrum amplitude modulation method for rolling bearing fault frequency determination

Abstract Signal preprocessing and feature extraction are decisive factors in determining the frequency of bearing faults. The presence of noise interference in the status signal of rolling bearings often hampers accurate fault detection. Although there are various methods for preprocessing vibration signals in rolling bearings, they need further improvement in terms of enhancing fault feature expression and localizing fault frequency bands. This limitation significantly hinders the accuracy of fault frequency determination. In order to enhance the representation of fault information on the frequency spectrum, this study proposes a combined approach that incorporates sparse stacked autoencoder (SSAE), wavelet packet decomposition (WPD), and adaptive spectrum amplitude modulation (ASAM). The resulting method is referred to as SSAE-WPD-ASAM. Firstly, the bearing vibration signal is decomposed by wavelet packet according to the scale and frequency band of the signal. On this basis, the signal reconstruction is realized based on the wavelet packet coefficient and energy distribution in different frequency bands. Secondly, for the whole life cycle signal, the reconstructed signal is self-encoded by sparse stacked autoencoder to achieve dimensionality reduction of the reconstructed signal. Then, the spare reconstructed signal is subjected to ASAM. Finally, through envelope demodulation, peak detection of fault frequency and empirical fault frequency comparison, the specific fault types of rolling bearings are determined. The proposed method is verified by theoretical simulation and three groups of practical experiments. The results show that the proposed method has a significant improvement in diagnostic efficiency and accuracy compared with traditional diagnostic methods.

Cutting state recognition of spiral bevel gear face hobbing

Abstract To investigate the issue of chatter vibration during the face hobbing process of spiral bevel gears, this study conducted experiments on the YKA2260 gear hobbing machine, collecting 142 sets of vibration acceleration signals. The optimal combination for wavelet threshold denoising was determined based on the signal-to-noise ratio (SNR), and preprocessing of the raw signals was performed. Utilizing wavelet packet decomposition, the energy entropy of wavelet packet coefficients was extracted as feature values. A recognition model for the cutting state during the face hobbing process of spiral bevel gears was established using a Support Vector Machine, achieving a recognition accuracy of up to 98.0769%. Furthermore, the model’s high recognition accuracy for small training samples in engineering applications was validated.