Waveform Components Research Articles

Numerical Modeling on Ocean-Bottom Seismograph P-Wave Receiver Function to Analyze Influences of Seawater and Sedimentary Layers

It is challenging to apply the receiver function method to teleseisms recorded by ocean-bottom seismographs (OBSs) due to a specific working environment that differs from land stations. Teleseismic incident waveforms reaching the area beneath stations are affected by multiple reflections generated by seawater and sediments and noise resulting from currents. Furthermore, inadequate coupling between OBSs and the seabed basement and the poor fidelity of OBSs reduce the signal-to-noise ratio (SNR) of seismograms, leading to the poor quality of extracted receiver functions or even the wrong deconvolution results. For instance, the poor results cause strong ambiguities regarding the Moho depth. This study uses numerical modeling to analyze the influences of multiple reflections generated by seawater and sediments on H-kappa stacking and the neighborhood algorithm. Numerical modeling shows that seawater multiple reflections are mixed with the coda waves of the direct P-wave and slightly impact the extracted receiver functions and can thus be ignored in subsequent inversion processing. However, synthetic seismograms have strong responses to the sediments. Compared to the waveforms of horizontal and vertical components, the sedimentary responses are too strong to identify the converted waves clearly. The extracted receiver functions correspond to the above influences, resulting in divergent results of H-kappa stacking (i.e., the Moho depth and crustal average VP/VS ratio are unstable and have great uncertainties). Fortunately, waveform inversion approaches (e.g., the neighborhood algorithm) are available and valid for obtaining the S-wave velocity structure of the crust–upper mantle beneath the station, with sediments varying in thickness and velocity.

Optimal control of a doubly fed induction generator of a wind turbine in cooperation with weak and rigid grids

In this article a control method for a rotor side converter (RSC) of a doubly fed induction generator of a wind turbine is developed. The doubly-fed induction generator (DFIG) system of a wind energy power plant that improves grid symmetrization was applied. The issue of optimal control was treated as an extended linear quadratic regulator (ELQR) having an extra set of exogenous inputs, which are source voltages of the electric grid. No additional knowledge of equations modelling the exogenous inputs was assumed. The proposed method is much more efficient than the currently available linear quadratic control methods. The control objective for a weak grid was to maintain the given value of the voltage module on load terminals and the given value of active power transferred from the stator’s winding. First harmonic components of relevant waveforms were used for this purpose. This task also required the DFIG system to provide reactive power to the grid. In the case of a rigid grid, this reactive power would be too high. Therefore, in this case, it was assumed that the system would supply only part of the required active and reactive power, based on its capabilities. It was required that the voltage and current ratings of the system, mainly the DFIG, were not exceeded. Therefore, the parameters of the network in these difficult failure cases were corrected only partially. The behaviour of the grid in the conditions of failure, as well as the return to the steady state after failure disappearance, were studied.

Model based fitting of pattern reversal visually evoked potentials provides a reliable characterization of waveform components

ObjectiveTo introduce a novel approach to analyzing pattern reversal visual evoked potentials (prVEPs) using a difference-of-gammas model-based fitting method. MethodsprVEP was recorded from uninjured youth ages 11–19 years during pre- or post-season sports evaluation. A difference-of-gammas model fit was used to extract the amplitude, peak time, and peak width of each of four gamma components. The within session reliability and stability of fits across a 6-month period were determined. To demonstrate an application of this analysis, changes in parameters across age were determined. ResultsA difference-of-gammas model consisting of four gamma functions was fit to the prVEP of 151 youth. Peak times and amplitudes of functions corresponded to standard measures of the N75, P100, and N135 components respectively, and a late gamma peak (mean peak time 219 ms). We extracted the peak width, which increased with each subsequent temporal peak. Parameter fits were reliable within sessions (correlation coefficient > 0.92 for all measured parameters; good agreement on Bland-Altman calculation) and were stable between sessions separated by less than 6 months (correlation coefficient > 0.90). Standard peak analysis metrics extracted from the difference-of-gamma model fits were largely consistent with gold-standard peak analysis measurements. ConclusionsThe difference-of-gammas model provides a stable and reliable within-participant representation of the global temporal variability of prVEP waveforms across a large sample of youth. SignificanceUsing difference-of-gammas model to characterize the global temporal variability of the prVEP waveform offers a promising direction to enhance analysis for identifying and following subtle changes in neurologic conditions.

ECG signal decomposition using Fourier analysis

This paper explores the Fourier decomposition method to approximate the decomposition of electrocardiogram (ECG) signals into their component waveforms, such as the QRS-complex and T-wave. We compute expansion coefficients using the ℓ1\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\ell _1$$\\end{document} Fourier transform and the traditional ℓ2\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\ell _2$$\\end{document} Fourier transform. Numerical examples are presented, and the analysis focuses on ECG signals as a real-world application, comparing the performance of the ℓ1\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\ell _1$$\\end{document} and ℓ2\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\ell _2$$\\end{document} Fourier transforms. Our results demonstrate that the ℓ1\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\ell _1$$\\end{document} Fourier transform significantly enhances the separation of ECG signal components, such as the QRS-complex and T-wave. This improvement is attributed to a notable reduction in the Gibbs phenomenon introduced by the Fourier-series expansion when using the ℓ1\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\ell _1$$\\end{document} Fourier transform, as opposed to the traditional ℓ2\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\ell _2$$\\end{document} Fourier transform.

Delineation of ECG waveform components using encoder-decoder architecture with Postprocess algorithm

Delineation of ECG waveform components using encoder-decoder architecture with Postprocess algorithm

Auditory brainstem response audiometry in tauopathy mouse model of human Alzheimer's disease.

Alzheimer's disease (AD) is a progressive neurodegenerative disorder in which changes in hearing sensitivity precede cognitive decline. Despite a well-known link between dementia and hearing loss, few AD model mouse lines have hearing characterized. We screened for hearing loss using auditory brainstem responses (ABR) in young (3-4 months) and aging (9-10 months) mice with a P301S tauopathy (PS19 mice). Compared to wild types, aging PS19 mice did not show accelerated hearing loss but did show latency differences in centrally generated ABR waveform components. These results suggest that tauopathy causes mild central auditory dysfunction in the absence of overt hearing loss.

Yield estimation of North Korean underground nuclear tests using Lg-wave source spectra

In seismic nuclear monitoring, attenuation models are important prerequisites for reliably estimating the explosive yield in an uncalibrated region without the occurrence of standard events. The seismic moment obtained by fitting source spectra is related to the source energy. This approach is appropriate for estimating yield, as the attenuation effects on the propagation path can be accurately considered. In this study, we collected 2022 vertical component waveforms in and around the Korean Peninsula from May 2010 to May 2022 to construct a high-resolution broadband Lg-wave attenuation model and inverted the Lg-wave source excitation spectra of the nuclear explosion simultaneously with attenuation correction. We obtained the scalar seismic moments by fitting the theoretical source spectra based on the Brune (J. Geophys. Res., 1970, 75, 4997–5009) model. Under the given emplacement conditions and burial depths, the seismic moments can be used to estimate yields of the North Korean nuclear tests, which are 4.6, 8.5, 19.9, 20.9, 24.7, and 337.4 kt for six nuclear explosions that occurred between 2006 and 2017. Our results are consistent with those obtained from previous teleseismic observations.

Assessing pressure wave components for aortic stiffness monitoring through spectral regression learning.

The ageing process notably induces structural changes in the arterial system, primarily manifesting as increased aortic stiffness, a precursor to cardiovascular events. While wave separation analysis is a robust tool for decomposing the components of blood pressure waveform, its relationship with cardiovascular events, such as aortic stiffening, is incompletely understood. Furthermore, its applicability has been limited due to the need for concurrent measurements of pressure and flow. Our aim in this study addresses this gap by introducing a spectral regression learning method for pressure-only wave separation analysis. Leveraging data from the Framingham Heart Study (2640 individuals, 55% women), we evaluate the accuracy of pressure-only estimates, their interchangeability with a reference method based on ultrasound-derived flow waves, and their association with carotid-femoral pulse wave velocity (PWV). Method-derived estimates are strongly correlated with the reference ones for forward wave amplitude ( ), backward wave amplitude ( ), and reflection index ( ) and moderately correlated with a time delay between forward and backward waves ( ). The proposed pressure-only method shows interchangeability with the reference method through covariate analysis. Adjusting for age, sex, body size, mean blood pressure, and heart rate, the results suggest that both pressure-only and pressure-flow evaluations of wave separation parameters yield similar model performances for predicting carotid-femoral PWV, with forward wave amplitude being the only significant factor (P < 0.001; 95% confidence interval, 0.056-0.097). We propose an interchangeable pressure-only wave separation analysis method and demonstrate its clinical applicability in capturing aortic stiffening. The proposed method provides a valuable non-invasive tool for assessing cardiovascular health.

Seismic Imaging of Shallow Depths using Ambient Noise Tomography in the Hyderabad Region, Eastern Dharwar Craton, India

ABSTRACT Hyderabad and its surrounding region are situated within Eastern Dharwar Craton, primarily composed of Archean granite. We use ambient noise tomography to obtain the 3-D shallow subsurface shear seismic velocity structure of 150 x 200 km2 area around Hyderabad. We used continuous waveforms that were recorded by a digital seismic network with 10 three-component broadband seismographs from November 2020 to December 2021. Using the vertical component of waveforms, the Empirical Green’s functions (EGFs) have been estimated, and the group velocity dispersion curves in the time period range of 1.0–10 s have been extracted from these. We inverted these dispersion curves to estimate 3-D variation of shear wave velocity by using direct surface wave tomographic method. The results suggest that the subsurface velocity structure is heterogeneous. The shear wave velocity varies from 3.4 to 3.7 km/s in the shallow subsurface. The limited seismicity in this area is mostly confined to the boundary between high and low-velocity changes.

A high-precision fusion bathymetry of multi-channel waveform curvature for bathymetric LiDAR systems

Airborne LiDAR bathymetry (ALB) system is an attractive and efficient method for nearshore bathymetry and underwater topography mapping. To ensure and improve the measurement accuracy and reliability in various water environments, different receivers using a segmented field of view (FOV) have been designed and are implemented in ALB. These are used to obtain various echo data of multiple channels and perform bathymetry at various water depths. However, in echo waveform processing, detailed information about the water body described by the different echo waveforms is lacking, making it difficult to overcome severe waveform superposition in extremely shallow areas or to separate the weak echo of the water bottom from the noise in extremely deep areas. Therefore, we employed a novel method of multi-channel waveform fusion bathymetry (MWFB) to achieve information complementarity, gain multi-channel waveforms, and detect the robustness of water and bottom echo signals. This method uses an adaptive signal waveform extraction of a multichannel fusion mechanism to eliminate systematic and random noise, and the waveform curvature is used to reconstruct the energy–waveform curves. Iterative decomposition based on waveform curvature and energy curves was conducted to improve the correctness of waveform decomposition and the reliability of waveform components. Furthermore, the water surface and bottom peaks were detected based on the multichannel waveform curvature, which enabled acquisition of the accurate temporal positions of the peaks and achieve high-accuracy bathymetry. According to the experimental results and a comparison with reference datasets, the highest bathymetric accuracy of the MWFB method reached an RMSE of 0.22. For the different study areas, the bathymetric accuracy of the slope reached 95%, 92%, and 97% at Ganquan Island, Lingyang Reef, and Bei Island, respectively. Furthermore, the bathymetric point numbers were effectively increased and bathymetric accuracy was also improved in shallow and deep water, which illustrates the superior bathymetric performance of the MWFB method and its ability to provide a high-efficiency waveform dataset through multi-channel data fusion. The novel LiDAR bathymetric method proposed in this study can effectively achieve high-accuracy nearshore bathymetry and seafloor topographical mapping.

Elimination of seismic characteristics of solid-filled in ultra-deep fractured-vuggy reservoirs

A large number of high-quality fractured-vuggy reservoirs (FVRs) are buried in ultra-deep (6500–9000 m) carbonate layers in the Tarim Basin, represented by the strong “beaded” reflection (SBR) on seismic profiles. This characteristic corresponds to the vast majority of high yield oil and gas wells, but there have also been a number of instances where the reservoir was filled with solid, leading to exploration failure. To avoid drilling failure, it is crucial to accurately identify fake reservoirs using geophysical methods. This paper proposes a novel method to eliminate the seismic characteristics of solid-filled in FVRs. Based on the waveform component decomposition (WCD) data, combined with drilling and logging data, an adaptive component reconstruction (ACR) can be completed to distinguish whether the reservoir is filled with solid or not, achieving the goal of eliminating the seismic characteristics of filling. Firstly, supervised WCD is performed on the target seismic data. Then, the waveform characteristics, which indicate the difference in the FVRs filled with solids, are combined accordingly with the sensitive attribute characteristics (instantaneous amplitude, heterogeneous anomalies, arc length, etc.) of each component data, and the feature templates of filled and non-filled FVRs are defined. Finally, the information entropy method is applied to optimise the component set that can distinguish the solid filled and non-filled FVRs and complete the component reconstruction. Forward modelling shows that this method can effectively distinguish the characteristics of FVRs that have been filled by solids from those that not. This method has achieved favourable results in the M block in the Tarim Basin. Successfully removed some or all the characteristic information of solid filled reservoirs confirmed by drilling from seismic data. The effectiveness of the method in identifying solid filled reservoirs has been confirmed by drilling, and the prediction rate of FVRs non-filled with solids has increased from 78% to 95%. This method can be used for reference in blocks with similar geological characteristics.

Solving variability: Accurately extracting feature components from ballistocardiograms.

A ballistocardiogram (BCG) is a vibration signal generated by the ejection of the blood in each cardiac cycle. The BCG has significant variability in amplitude, temporal aspects, and the deficiency of waveform components, attributed to individual differences, instantaneous heart rate, and the posture of the person being measured. This variability may make methods of extracting J-waves, the most distinct components of BCG less generalizable so that the J-waves could not be precisely localized, and further analysis is difficult. This study is dedicated to solving the variability of BCG to achieve accurate feature extraction. Inspired by the generation mechanism of the BCG, we proposed an original method based on a profile of second-order derivative of BCG waveform (2ndD-P) to capture the nature of vibration and solve the variability, thereby accurately localizing the components especially when the J-wave is not prominent. In this study, 51 recordings of resting state and 11 recordings of high-heart-rate from 24 participants were used to validate the algorithm. Each recording lasts about 3 min. For resting state data, the sensitivity and positive predictivity of proposed method are: 98.29% and 98.64%, respectively. For high-heart-rate data, the proposed method achieved a performance comparable to those of low-heart-rate: 97.14% and 99.01% for sensitivity and positive predictivity, respectively. Our proposed method can detect the peaks of the J-wave more accurately than conventional extraction methods, under the presence of different types of variability. Higher performance was achieved for BCG with non-prominent J-waves, in both low- and high-heart-rate cases.

Analyze the Difference Between Rotational and Translational Motions Produced by High-speed Train

The development of fiber optic gyroscope technology makes it possible to directly measure the rotational ground-motion. The joint observation of the three-component rotational seismometer and the traditional three-component translational seismometer is a trend of future. In this paper, the translational and rotational signals generated during the high-speed trainway are studied, and the three-component translational and rotational seismometers are fixed together to realize the joint recording of the high-speed trainway signal. Comparing the translational and rotational three-component data, we find that only one component waveforms and the spectrum have certain consistency. However, the difference between them indicates that the rotation and translation signals generated by the shallow surface wave signal have different frequency bands.

Poster Session I: Comparison of tripolar and traditional electrodes: Waveform morphology.

The waveforms of visual evoked potentials (VEPs) recorded by tripolar electrodes are different from those recorded by traditional electrode arrays. Traditional arrays record potentials using both an active and a reference electrode. The location of this reference will affect the waveform amplitude and shape. Conversely, tripolar electrodes measure the surface Laplacian across three concentric rings housed within a single electrode surface. These differences may influence the morphology of evoked responses. We compared these two modalities by recording pattern-reversal VEPs to two sizes of checkerboard. Visual inspection of the VEPs suggest that the signals recorded by the tripolar system have attenuated higher frequencies and increased latency of the major waveform components. Root-mean-square comparison of the two signal types confirm attenuation at the higher frequencies in the tripolar recording. Additionally, there is a cumulative delay present within both the large and small check conditions, such that each subsequent component recorded by the tripolar electrodes is shifted increasingly later in time compared to the same components recoded by the traditional electrodes. Such latency shifts may be indicative of a difference in the physiological sources that are measured by the two EEG systems.

SwinDAE: Electrocardiogram Quality Assessment Using 1D Swin Transformer and Denoising AutoEncoder.

Electrocardiogram (ECG) signals have wide-ranging applications in various fields, and thus it is crucial to identify clean ECG signals under different sensors and collection scenarios. Despite the availability of a variety of deep learning algorithms for ECG quality assessment, these methods still lack generalization across different datasets, hindering their widespread use. In this paper, an effective model named Swin Denoising AutoEncoder (SwinDAE) is proposed. Specifically, SwinDAE uses a DAE as the basic architecture, and incorporates a 1D Swin Transformer during the feature learning stage of the encoder and decoder. SwinDAE was first pre-trained on the public PTB-XL dataset after data augmentation, with the supervision of signal reconstruction loss and quality assessment loss. Specially, the waveform component localization loss is proposed in this paper and used for joint supervision, guiding the model to learn key information of signals. The model was then fine-tuned on the finely annotated BUT QDB dataset for quality assessment. SwinDAE achieved 0.02-0.13 mean F1 score improvement on the BUT QDB dataset compared to multiple deep learning methods, and demonstrated applicability on two other datasets. The proposed SwinDAE shows strong generalization ability on different datasets, and surpasses other state-of-the-art deep learning methods on multiple evaluation metrics. In addition, the statistical analysis for SwinDAE prove the significance of the performance and the rationality of the prediction. SwinDAE can learn the commonality between high-quality ECG signals, exhibiting excellent performance in the application of cross-sensors and cross-collection scenarios.

Analysis of Swarm Earthquakes on the Territory of the Elbrus Volcanic Center

The paper presents the results of the analysis of swarm seismic events recorded in 2018 in the area of the Elbrus volcanic center by the North Caucasian seismological network of the FRC GS RAS. The standard method of station processing shows the hypocenters of recorded events in the area of the Elbrus volcanic center. The hypocenters of seismic events in swarm sequences were refined using the LOS software package, a complex combined location algorithm, methods of minimizing the time discrepancy in the source and grid search with depth enumeration. The described approach improves the reliability of location due to accounting for inaccuracies associated with incorrectly taken times of phase arrivals and minor deviations in travel times, depending on the velocity model of the medium used. According to the results of the study of the focal mechanisms of the strongest earthquakes, the type of movement was established, which is consistent with the kinematics of the Elbrus-Mineralnye Vody fault zone. Correlation analysis of the waveforms of the vertical component filtered in the frequency range of 1‒10 Hz, including P and S waves, for the nearest station shows a high coincidence of the analyzed data (correlation coefficient R ≥ 0.85), which indicates a high probability that events have similar sources and paths propagation of seismic waves. Comparison of the power spectral density of individual earthquakes in swarm sequences also demonstrates high similarity.

Superoscillations Deliver Superspectroscopy.

In ordinary circumstances the highest frequency present in a wave is the highest frequency in its Fourier decomposition. It is however possible for there to be a spatial or temporal region where the wave locally oscillates at a still greater frequency in a phenomenon known as superoscillation. Superoscillations find application in wide range of disciplines, but at present their generation is based upon constructive approaches that are difficult to implement. Here, we address this, exploiting the fact that superoscillations are a product of destructive interference to produce a prescription for generating superoscillations from the superposition of arbitrary waveforms. As a first test of the technique, we use it to combine four quasisinusoidal THz waveforms to produce THz optical superoscillations for the first time. The ability to generate superoscillations in this manner has potential application in a wide range of fields, which we demonstrate with a method we term "superspectroscopy." This employs the generated superoscillations to obtain an observed enhancement of almost an order of magnitude in the spectroscopic sensitivity to materials whose resonance lies outside the range of the component waveform frequencies.

New pilot protection method based on the current fault component waveform similarity

AbstractIn order to improve the ability of the modular multi‐level converter to quickly clear and isolate faults in the multi‐terminal flexible DC grid, this paper analyzes the characteristics of current faults in the multi‐terminal flexible direct current grid, introduces the comprehensive distance of current fault waveform, and proposes a line protection method based on the similarity of current fault waveform. This protection method firstly uses Euclid Dynamic Time Warping (DTW) distance and entropy weight method to process electrical quantity signals, calculates the comprehensive distance of current fault waveform of two converter stations, and builds the protection criterion based on the similarity of current fault waveform. Then, the fault identification is realized by using the calculated comprehensive distance, and the fault pole identification is realized by using the ratio of comprehensive distance between positive and negative poles. Finally, PSCAD was used to build a simulation model to output the fault data under different fault conditions, and MATLAB was used to process the fault data to verify the protection algorithm. The simulation results show that the proposed method has good resistance to high resistance, and its rapidity and adaptability are further improved. Compared with other similarity methods, this method combines Euclidean DTW distance and entropy method to process the signal, realizes the fault pole identification, and can still operate correctly under the interference of 500 Ω transition resistance and 20 dB noise ratio. The research results will provide a theoretical basis for the safe and reliable operation of flexible DC power grid.

Detailed muscle state analysis method based on real-time wavelet analysis of surface myoelectric potential

Myoelectric refers to the use of electricity generated by muscles and is harnessed in the development of electrically powered prostheses, which are controlled by electromyographic (EMG) signals created in the residual musculature. Assistant professor Hidetoshi Nagai, Department of Artificial Intelligence, Kyushu Institute of Technology, Japan, is interested in surface myoelectric signals and is working on a project to develop technology that can advance their use. Nagai will capture motor unit activities using surface electromyography, which is easy to measure during exercise, and use this as the basis for more detailed muscle activity analysis. The methods Nagai has developed require no special equipment, other than the ability to sample at frequencies of several tens of kHz, and only require a single channel, which indicates the potential for more sophisticated analysis when multiple channels of information are present. It is also a simple and lightweight process that can be executed in real time. Conventional analysis and evaluation of muscle activity cannot be performed from the perspective of motor unit activity but Nagai has built on the basic premise that given that muscle activity is the sum of motor unit activities, the analysis of muscle activity should be based on the analysis of motor unit activity. He will analyse the surface EMG signal from the viewpoint of its component waveform the motor unit waveform so that muscle activity analysis can be performed as it should be.

An interpretable multi-scale lightweight network for patient-ventilator asynchrony detection during mechanical ventilation

Patient-ventilator asynchrony (PVA) is commonly the result of an inconsistency between the patient's respiratory rhythm and the ventilator rate during mechanical ventilation, which will lead to serious prognostic problems. Therefore, it is necessary to develop a reliable automatic recognition algorithm. However, many current intelligent algorithms, suffer from poor detection efficiency due to the limited datasets, lack of interpretability, with huge amounts of parameters, which are difficult to deploy on resource-constrained hardware devices. In this paper, an interpretable multi-scale lightweight network, Mobiformer, is proposed for automatic PVA identification. The model adopts a novel lightweight parallel structure, combining the ability of parameter sharing of convolutional network and the global interaction of self-attention mechanism for multi-level feature extraction. It can dynamically capture the global dependencies while extracting the local information of the waveforms, thus detecting the occurrence of PVA more accurately. In addition, a relative position encoding method is employed and a visualization module is built to improve the classification accuracy while making the result interpretable, avoiding the black-box problem. In the test, the identification accuracy, sensitivity, specificity, and F1 scores of the model are 0.987, 0.983, 0.986, and 0.989, respectively. Moreover, the critical waveform components of different categories are highlighted, consistent with the understanding of PVA by physicians. The experimental results show that our model can identify PVA more effectively and make the decision interpretable, which offers tremendous potential for future clinical applications.