Application of sparse S transform network with knowledge distillation in seismic attenuation delineation

Time-frequency analysis is a successfully used tool for analyzing the local features of seismic data.However,it suffers from several inevitable limitations,such as the restricted time-frequency resolution,the difficulty in selecting parameters,and the low computational efficiency.Inspired by deep learning,we suggest a deep learning-based workflow for seismic time-frequency analysis.The sparse S transform network(SSTNet)is first built to map the relationship between synthetic traces and sparse S transform spectra,which can be easily pre-trained by using synthetic traces and training labels.Next,we introduce knowledge distillation(KD)based transfer learning to re-train SSTNet by using a field data set without training labels,which is named the sparse S transform network with knowledge distillation(KD-SSTNet).In this way,we can effectively calculate the sparse time-frequency spectra of field data and avoid the use of field training labels.To test the availability of the suggested KD-SSTNet,we apply it to field data to estimate seismic attenuation for reservoir characterization and make detailed comparisons with the traditional time-frequency analysis methods.

Amplitude spectrum compensation and phase spectrum correction of seismic data based on the generalized S transform

We propose a method for the compensation and phase correction of the amplitude spectrum based on the generalized S transform. The compensation of the amplitude spectrum within a reliable frequency range of the seismic record is performed in the S domain to restore the amplitude spectrum of reflection. We use spectral simulation methods to fit the time-dependent amplitude spectrum and compensate for the amplitude attenuation owing to absorption. We use phase scanning to select the time-, space-, and frequencydependent phases correction based on the parsimony criterion and eliminate the residual phase effect of the wavelet in the S domain. The method does not directly calculate the Q value; thus, it can be applied to the case of variable Q. The comparison of the theory model and field data verify that the proposed method can recover the amplitude spectrum of the strata reflectivity, while eliminating the effect of the residual phase of the wavelet. Thus, the wavelet approaches the zero-phase wavelet and, the seismic resolution is improved.

Ground roll attenuation using a time-frequency dependent polarization filter based on the S transform

The ground roll and body wave usually show significant differences in arrival time, frequency content, and polarization characteristics, and conventional polarization filters that operate in either the time or frequency domain cannot consider all these elements. Therefore, we have developed a time-frequency dependent polarization filter based on the S transform to attenuate the ground roll in seismic records. Our approach adopts the complex coefficients of the S transform of the multi-component seismic data to estimate the local polarization attributes and utilizes the estimated attributes to construct the filter function. In this study, we select the S transform to design this polarization filter because its scalable window length can ensure the same number of cycles of a Fourier sinusoid, thereby rendering more precise estimation of local polarization attributes. The results of applying our approach in synthetic and real data examples demonstrate that the proposed polarization filter can effectively attenuate the ground roll and successfully preserve the body wave.

Fractional S transform-Part 1:Theory

The S transform, which is a time-frequency representation known for its local spectral phase properties in signal processing, uniquely combines elements of wavelet transforms and the short-time Fourier transform （STFT）. The fractional Fourier transform is a tool for non-stationary signal analysis. In this paper, we define the concept of the fractional S transform （FRST） of a signal, based on the idea of the fractional Fourier transform （FRFT） and S transform （ST）, extend the S transform to the time-fractional frequency domain from the time- frequency domain to obtain the inverse transform, and study the FRST mathematical properties. The FRST, which has the advantages of FRFT and ST, can enhance the ST flexibility to process signals. Compared to the S transform, the FRST can effectively improve the signal time- frequency resolution capacity. Simulation results show that the proposed method is effective.

High-frequency compensation for seismic data based on adaptive generalized S transform

The low-pass fi ltering eff ect of the Earth results in the absorption and attenuation of the high-frequency components of seismic signals by the stratum during propagation.Hence,seismic data have low resolution.Considering the limitations of traditional high-frequency compensation methods,this paper presents a new method based on adaptive generalized S transform.This method is based on the study of frequency spectrum attenuation law of seismic signals,and the Gauss window function of adaptive generalized S transform is used to fi t the attenuation trend of seismic signals to seek the optimal Gauss window function.The amplitude spectrum compensation function constructed using the optimal Gauss window function is used to modify the time-frequency spectrum of the adaptive generalized S transform of seismic signals and reconstruct seismic signals to compensate for high-frequency attenuation.Practical data processing results show that the method can compensate for the high-frequency components that are absorbed and attenuated by the stratum,thereby eff ectively improving the resolution and quality of seismic data.

Novel robust S transform based on the clipping method

This paper presents a novel robust S transform algorithm based on the clipping method to process signals corrupted by impulsive noise.The proposed algorithm is introduced to determine the clipping threshold value according to the characteristics of the signal samples.Signals in various impulsive noise models are considered to illustrate that the robust S transform can achieve better performance than the standard S transform.Moreover,mean square errors for instantaneous frequency estimation of the robust S transform are compared with that of the standard S transform,showing that the robust S transform can achieve significantly improved instantaneous frequency estimation for the signals in impulsive noise.

基于S-YOLO V5和Vision Transformer的视频内容描述算法

视频内容描述的自动生成是结合计算机视觉和自然语言处理等相关技术提出的一种新型交叉学习任务。针对当前视频内容生成描述模型可读性不佳的问题,本研究提出一种基于S-YOLO V5和Vison Transformer(ViT)的视频内容描述算法。首先,基于神经网络模型KATNA提取关键帧,以最少帧数进行模型训练;其次,利用S-YOLO V5模型提取视频帧中的语义信息,并结合预训练ResNet101模型和预训练C3D模型提取视频静态视觉特征和动态视觉特征,并对两种模态特征进行融合;然后,基于ViT结构的强大长距离编码能力,构建模型编码器对融合特征进行长距离依赖编码;最后,将编码器的输出作为LSTM解码器的输入,依次输出预测词,生成最终的自然语言描述。通过在MSR-VTT数据集上进行测试,本研究模型的BLEU-4、METEOR、ROUGEL和CIDEr分别为42.9、28.8、62.4和51.4;在MSVD数据集上进行测试,本研究模型的BLEU-4、METEOR、ROUGEL和CIDEr分别为56.8、37.6、74.5以及98.5。与当前主流模型相比,本研究模型在多项评价指标上表现优异。

K-S变换及其电网超谐波时频分析应用

依据FFT→优化窗→IFFT思路,突破线性时频变换的窗函数积分性能桎梏,实现高性能优化窗函数的线性时频变换应用,建立新型时频变换算法——K-S变换.对信号x(t)的FFT频谱向量进行频移处理后,与该频移点下Kaiser优化窗的频谱向量进行Hadamard乘积,再将乘积结果进行FFT逆变换(IFFT),构造出K-S变换复时频矩阵,由此获得x(t)的时间-频率-幅值、时间-频率-相位三维信息;给出逆变换的数学推导与局部性质、线性性质和变分辨率特性;0~150 kHz电网的稳态与时变超谐波信号仿真实验表明,K-S变换的时域、频域分辨能力均优于流行的短时傅里叶变换、S变换,具有优良的变分辨率性能;0~40 kHz超谐波信号的实测证明,基于K-S变换的超谐波电压幅值测量绝对误差均小于0.032 3 V.

基于哨兵函数和S变换的风力机叶片材料损伤特性研究

利用声发射检测技术研究玻璃纤维增强环氧树脂复合材料的损伤特性,在此过程中,采用哨兵函数来表征该材料的损伤程度,并通过S变换和模糊C均值(FCM)聚类来分析声发射信号,从而获得材料的损伤特征。三点弯曲实验结束后对试件断口进行扫描电子显微镜(SEM)拍照来验证,可得:通过对SEM照片的分析得到基体开裂、纤维脱粘、分层破坏、纤维断裂4种损伤模式;对整个声发射事件进行哨兵函数分析,观察到试件在弯曲过程中哨兵函数曲线呈明显下降趋势;对依据哨兵函数划分的不同阶段的信号进行VMD降噪处理,然后采用S变换进行时频分析得到不同损伤的特征频率,最后采用FCM聚类进行验证。结果表明:哨兵函数值的突变可作为材料断裂的预警信号,材料损伤类型的识别可依据S变换的频率分布结果进行确定。

基于S变换双阈值法的汽车零部件载荷谱加速编辑

针对S变换编辑法加速编辑效果不佳的问题,提出了一种基于S变换双阈值编辑方法。该方法先对载荷谱进行S变换以获取最大幅值谱,在以双阈值识别并保留的幅值谱片段为依据保留对应的载荷谱片段后,再将其拼接成加速后的载荷谱,最后对比分析S变换双阈值编辑法与S变换编辑法编辑的加速谱的统计参数、功率谱密度、穿级计数和疲劳仿真结果。研究结果表明,S变换双阈值编辑法可明显压缩原始载荷时间,且其压缩效率高于S变换编辑法,转向节的疲劳损伤和寿命分析误差更小,适用于汽车零部件载荷谱加速编辑研究。

基于S变换模能量分析的直流电缆故障测距

针对现有高压直流电缆故障定位方法故障定位精度低、定位模型泛化能力差等问题,文中在考虑电缆故障后故障点产生的暂态信号中含有丰富的故障信息的特征,提出一种时频域能量分析的故障测距方法。该方法利用多分类支持向量机(SVM)的特性构建直流电缆故障定位模型,利用直流电缆相模解耦矩阵对电缆电气量解耦,并采用S变换进行时频域变换,将得到的模量构造模能量,将其作为样本对SVM模型训练。实验结果表明,所提方法对于不同故障类型在不同过渡电阻及不同位置故障下均可进行准确测距。

经验小波变换和改进S变换结合的电能质量检测与识别方法

为分析不确定干扰因素影响下的实际电力网络电能质量问题,提出一种经验小波变换(EWT)和改进S变换相结合的电能质量检测与识别方法。该方法一方面利用EWT联合归一化直接正交(NDQ)算法和奇异值分解(SVD)算法准确提取调幅-调频分量的频率、幅值和时间参数,另一方面考虑到EWT算法在高噪声环境下瞬时幅值波动的问题,引入改进S变换提取高噪声干扰下的电能质量扰动时频信息,最后,基于EWT和改进S变换提取的扰动特征向量,利用基于改进粒子群优化算法(IPSO)优化支持向量机(SVM)的电能质量扰动识别分类器实现扰动类型的精确识别。仿真和实验表明所提方法在复合扰动识别分类时平均识别准确率为93.23%,且能够准确识别4种实测扰动信号。

S-100β CysC和NF-κB对急性缺血性脑卒中患者静脉溶栓后出血转化的预测价值

目的 探讨急性缺血性脑卒中(AIS)外周血中S-100β、CysC和NF-κB水平对静脉溶栓后出血转化的影响及预测价值。方法 收集2019-03—2022-03接受溶栓治疗的AIS患者140例,根据溶栓后24 h是否发生出血转化(HT)将患者分为非HT组(n=112)和HT组(n=28)。比较2组一般临床资料,采用多因Logistic回归分析影响HT发生的危险因素,受试者工作特征(ROC)曲线评估S-100β、CysC和NF-κB预测HT发生的价值。结果 HT组与非HT组相比,患者年龄、发病至溶栓时间、房颤、TOAST分型、C反应蛋白、凝血酶原时间、S-100β、CysC、NF-κB、白质高信号和脑微出血等均有统计学差异(P<0.05)。Logistic多因素回归分析显示,房颤、S-100β、CysC和NF-κB为影响HT发生的危险因素。S-100β、CysC和NF-κB预测AIS患者静脉溶栓后出血转化的曲线下面积分别为0.915(0.902~0.923)、0.874(0.856~0.882)和0.789(0.771~0.796),均具有一定的预测价值。结论 S-100β、CysC和NF-κB为AIS患者静脉溶栓后HT发生的危险因素,对HT发生具有一定的预测价值。

基于改进混沌麻雀搜索算法S盒与神经元电路的图像加密技术

鉴于目前的图像加密算法所采用的S盒结构过于简单,易被破解。选用由改进的脉冲耦合尖峰神经元电路得到的混沌映射(MPSNC),并使用该映射对混沌麻雀搜索算法(CSSA)进行优化。随后,结合MPSNC映射和改进后的CSSA,设计新型的混沌S盒。最后,基于MPSNC映射和新型混沌S盒构造一种图像加密算法,为增加S盒的变化,对其进行S8变换。通过迭代混沌映射生成混沌序列,使用混沌序列和S8S盒对图像进行扩散和置乱操作,得到加密后的图像。经实验验证,该加密算法参数丰富,复杂度高,并由于S盒的应用,提高了算法的非线性度,具有良好的性能,适用于图像加密。

基于同步压缩广义S变换的电力系统次/超同步振荡检测

为实现电力系统次/超同步振荡的快速、准确辨识,提出了一种基于同步压缩广义S变换(synchrosqueezing generalized S transform, SSGST)和改进稀疏时域法(improved sparse time domain method,ISTD)结合的次/超同步振荡辨识方法。该方法首先利用能量比函数对电力系统广域量测信息实时检测,当检测到信号能量发生突变时,利用SSGST对检测到的振荡信号分解得到相应的SSGST时频系数矩阵;然后通过改进的脊线提取方法在时频域实现对各振荡分量的最优轨迹搜索;进一步,结合最优轨迹时频索引重构各振荡分量的时域分量,并利用ISTD辨识方法计算出各振荡分量的频率和阻尼比系数;最后,通过自合成模拟信号、双馈风电场经串补并网系统仿真信号和某实际风电场实测数据验证了所提方法的准确性和有效性。

频率矫正广义S变换及在储层含油气性预测中的应用

随着我国油气勘探程度的不断加深,陆相页岩油气等非常规油气逐渐成为我国勘探开发的重点,陆相页岩非均质性强、储层薄且纵横向变化剧烈等典型特征导致含油气性预测精度差,影响后续井位部署和生产等.时频变换方法是储层含油气性预测最有效的工具.为了提高储层含油气性预测的精度,本文提出了一种频率矫正广义S变换(Frequency-Corrected Generalized S-Transform, FCGST),该变换既具有S变换多尺度多分辨率的特性,又可以通过自适应选取窗函数参数以控制窗函数的带宽和主频等,这一特性可以适用于不同工区地震数据的处理与解释.与传统的S变换和广义S变换不同,该变换修正了传统S变换和广义S变换中的尺度因子,改善了S变换和广义S变换等传统方法中时频谱主频向高频方向偏移的问题,有利于精确描述复杂储层含油气性.合成数据算例表明,频率矫正广义S变换提升了地震信号局部时频特征的表征能力;将频率矫正广义S变换应用于鄂尔多斯盆地某工区储层含油气性预测中,实钻结果数据进一步验证了该时频分析方法在储层含油气性解释中的有效性和潜力.

基于同步提取广义S变换的机械故障诊断方法研究

现有的同步提取变换(synchroextracting transform, SET)窗函数固定缺乏灵活性,在进行故障诊断时很难有效获取到高时频精度和高抗干扰性能的瞬时频率,针对此问题,结合广义S变换可以自适应调节窗函数宽度的优点,提出一种基于同步提取广义S变换(synchroextracting generalized Stransform, SEGST)的机械故障诊断方法。SEGST方法的特点在于将Rényi熵作为度量时频聚集性的标准,通过在高斯窗函数中引入2个尺度调节因子来选择参数的最佳值,对得到的广义S变换二维时频谱构造出同步提取算子来提取时频脊线处的时频系数,该算子能保留与信号的时变特征最相关的TF信息,剔除多余的模糊时频能量,从而得到高时频分辨率的时频能量特征。仿真结果表明,所提方法不论在时频分辨率方面,还是在噪声鲁棒性方面,都优于传统时频分析方法,并且保持了良好的重构性。最后,将所提方法应用于航空发动机高速滚动轴承故障诊断中,结果表明,该方法能够准确识别故障信号中的特征频率。

同步挤压S变换与τ⁃p变换联合的微地震信号消噪方法

微地震监测是指导页岩气开采水力压裂作业和评价压裂效果的常用手段。地面监测所采集的微地震信号能量弱、信噪比低,微地震事件识别困难,严重影响定位的准确性。针对低信噪比地面微地震监测资料,联合使用同步挤压S变换、谱分解和τ⁃p变换,提出一种新的消噪方法。首先对监测资料进行时差校正,将微地震信号的同相轴校平;之后使用同步挤压S变换对校平后的资料进行谱分解获取单频切片;再对每个单频切片进行τ⁃p变换,并根据τ⁃p变换的结果获取微地震信号位置;最后根据信号的位置在时频域完成消噪。实际低信噪比地面微地震监测数据的处理结果表明,新方法可以获得理想的消噪结果。

基于S变换的临汾水平摆倾斜观测数据年变信息提取及预测指标确定

临汾台水平摆多年来监测能力及资料的映震能力备受关注,为推进长周期破年变分析在临汾台水平摆资料的应用,本文利用一种基于S变换时频分析方法的破年变信息分离提取方法,采用双向非对称阈值策略,结合R值评分及Molchan图表法,构建了临汾水平摆东西测项破年变预测指标,给出了长周期信号变化(ANA)和短周期信号变化(ONA)对于不同震级档的最佳阈值、最佳预报范围、最佳预报时窗,预报效能R值均大于R0值,可信度高。其中ANA指标对于500 km以内的ML≥5级的预报效能最好。

基于同步提取增强广义S变换的柴油机气门性能退化状态评估

柴油机在运行过程中气门间隙逐渐增大,其状态会随气门性能退化而发生改变。针对传统状态评估方法难以对其气门性能退化状态进行准确评估的问题,提出基于同步提取增强广义S变换(Synchro Extracting Enhanced Generalized S-Transform,SEEGST)的柴油机气门性能退化状态评估方法。通过传感器采集反映柴油机状态的振动信号;为解决传统信号时频分析方法存在时频分辨率低、能量聚集性弱等问题,基于同步提取算法与广义S变换提出SEEGST时频分析方法,将振动信号转换为二维时频图;利用MLP-Mixer模型提取时频图像特征进行训练,实现柴油机状态评估。通过柴油机状态监测实验台开展气门性能退化实验,将所提方法与SSGST-MLPMixer、GST-MLPMixer、SEEGST-ViT、SEEGST-2DCNN、FFT spectrum-1DCNN 5种传统方法对比。实验结果表明:所提方法的整体评估准确率达到98.96%,可有效应用于柴油机气门性能退化状态评估领域,为开展柴油机气门性能退化状态评估提供一种新的思路。