Time-Lapse Full-Waveform Inversion Using Cross-Correlation-Based Dynamic Time Warping

Offshore carbon capture, utilization, and storage(OCCUS) is regarded as a crucial technology for mitigating greenhouse gas emissions.Quantitative monitoring maps of sealed carbon dioxide are necessary in a comprehensive OCCUS project. A potential high-resolution method for the aforementioned purpose lies in the full-waveform inversion(FWI) of time-lapse seismic data. However, practical applications of FWI are severely restricted by the well-known cycle-skipping problem. A new time-lapse FWI method using cross-correlation-based dynamic time warping(CDTW) is proposed to detect changes in the subsurface property due to carbon dioxide(CO_(2)) injection and address the aforementioned issue. The proposed method, namely CDTW, which combines the advantages of cross-correlation and dynamic time warping, is employed in the automatic estimation of the discrepancy between the seismic signals simulated using the baseline/initial model and those acquired. The proposed FWI method can then back-project the estimated discrepancy to the subsurface space domain, thereby facilitating retrieval of the induced subsurface property change by taking the difference between the inverted baseline and monitor models. Numerical results on pairs of signals prove that CDTW can obtain reliable shifts under amplitude modulation and noise contamination conditions. The performance of CDTW substantially outperforms that of the conventional dynamic time warping method. The proposed time-lapse fullwaveform inversion(FWI) method is applied to the Frio-2 CO_(2) storage model. The baseline and monitor models are inverted from the corresponding time-lapse seismic data. The changes in velocity due to CO_(2) injection are reconstructed by the difference between the baseline and the monitor models.

Method of Time Series Similarity Measurement Based on Dynamic Time Warping

With the rapid development of mobile communication all over the world,the similarity of mobile phone communication data has received widely attention due to its advantage for the construction of smart cities.Mobile phone communication data can be regarded as a type of time series and dynamic time warping(DTW)and derivative dynamic time warping(DDTW)are usually used to analyze the similarity of these data.However,many traditional methods only calculate the distance between time series while neglecting the shape characteristics of time series.In this paper,a novel hybrid method based on the combination of dynamic time warping and derivative dynamic time warping is proposed.The new method considers not only the distance between time series,but also the shape characteristics of time series.We demonstrated that our method can outperform DTW and DDTW through extensive experiments with respect to cophenetic correlation.

Partition of GB-InSAR deformation map based on dynamic time warping and k-means

Ground-based interferometric synthetic aperture radar(GB-InSAR)can take deformation measurement with a high accuracy.Partition of the GB-InSAR deformation map benefits analyzing the deformation state of the monitoring scene better.Existing partition methods rely on labelled datasets or single deformation feature,and they cannot be effectively utilized in GBInSAR applications.This paper proposes an improved partition method of the GB-InSAR deformation map based on dynamic time warping(DTW)and k-means.The DTW similarities between a reference point and all the measurement points are calculated based on their time-series deformations.Then the DTW similarity and cumulative deformation are taken as two partition features.With the k-means algorithm and the score based on multi evaluation indexes,a deformation map can be partitioned into an appropriate number of classes.Experimental datasets of West Copper Mine are processed to validate the effectiveness of the proposed method,whose measurement points are divided into seven classes with a score of 0.3151.

Grey incidence clustering method based on multidimensional dynamic time warping distance

The traditional grey incidence degree is mainly based on the distance analysis methods, which is measured by the displacement difference between corresponding points between sequences. When some data of sequences are missing (inconsistency in the length of the sequences), the only way is to delete the longer sequences or to fill the shorter sequences. Therefore, some uncertainty is introduced. To solve this problem, by introducing three-dimensional grey incidence degree (3D-GID), a novel GID based on the multidimensional dynamic time warping distance (MDDTW distance-GID) is proposed. On the basis of it, the corresponding grey incidence clustering (MDDTW distance-GIC) method is constructed. It not only has the simpler computation process, but also can be applied to the incidence comparison between uncertain multidimensional sequences directly. The experiment shows that MDDTW distance-GIC is more accurate when dealing with the uncertain sequences. Compared with the traditional GIC method, the precision of the MDDTW distance-GIC method has increased nearly 30%.

Hand Gesture Recognition by Accelerometer-Based Cluster Dynamic Time Warping

Aiming at the diversity of hand gesture traces by different people,the article presents novel method called cluster dynamic time warping( CDTW),which is based on the main axis classification and sample clustering of individuals. This method shows good performance on reducing the complexity of recognition and strong robustness of individuals. Data acquisition is implemented on a triaxial accelerometer with 100 Hz sampling frequency. A database of 2400 traces was created by ten subjects for the system testing and evaluation. The overall accuracy was found to be 98. 84% for user independent gesture recognition and 96. 7% for user dependent gesture recognition,higher than dynamic time warping( DTW),derivative DTW( DDTW) and piecewise DTW( PDTW) methods.Computation cost of CDTW in this project has been reduced 11 520 times compared with DTW.

基于PDES系统的Time Warp性能研究

随着分布式仿真规模的日益扩大,高性能并行计算技术的不断发展,并行与分布式仿真正逐渐成为新的研究热点。时间管理技术是决定并行与分布式仿真正确性和可重复性的关键技术,直接影响着仿真系统的整体性能。而TW乐观时间管理机制的许多思想和概念一直为现在各种算法所借鉴和沿用。文章建立了一个TW的性能分析模型,利用该模型对事件数量、平均回退长度、回退概率及其上限等进行了分析,分析结果对并行离散事件仿真系统的研发与改进具有一定的指导意义和参考价值。

Efficient Dynamic Time Warping by Adaptively Controlling the Valid Warping Range

Dynamic time warping(DTW)spends most of the time in generating the correlation table,and it establishes the global path constraints to reduce the time complexity.However,the global constraints restrain just in terms of the time axis.In this paper,we therefore propose another version of DTW,to be called branch-and-bound DTW(BnB-DTW),which adaptively controb its global path constraints by reflecting the contents of input patterns. Experimental results show that the suggested BnB-DTW algorithm performs more efficiently than other conventional DTW approaches while not increasing the optimal warping cost.

Merge-Weighted Dynamic Time Warping for Speech Recognition

Obtaining training material for rarely used English words and common given names from countries where English is not spoken is difficult due to excessive time, storage and cost factors. By considering personal privacy, language- independent （LI） with lightweight speaker-dependent （SD） automatic speech recognition （ASR） is a convenient option to solve tile problem. The dynamic time warping （DTW） algorithm is the state-of-the-art algorithm for small-footprint SD ASR for real-time applications with limited storage and small vocabularies. These applications include voice dialing on mobile devices, menu-driven recognition, and voice control on vehicles and robotics. However, traditional DTW has several lhnitations, such as high computational complexity, constraint induced coarse approximation, and inaccuracy problems. In this paper, we introduce the merge-weighted dynamic time warping （MWDTW） algorithm. This method defines a template confidence index for measuring the similarity between merged training data and testing data, while following the core DTW process. MWDTW is simple, efficient, and easy to implement. With extensive experiments on three representative SD speech recognition datasets, we demonstrate that our method outperforms DTW, DTW on merged speech data, the hidden Markov model （HMM） significantly, and is also six times faster than DTW overall.

An indoor fusion navigation algorithm using HV-derivative dynamic time warping and the chicken particle flter

The use of dead reckoning and fngerprint matching for navigation is a widespread technical method.However,fngerprint mismatching and low fusion accuracy are prevalent issues in indoor navigation systems.This work presents an improved dynamic time warping and a chicken particle flter to handle these two challenges.To generate the Horizontal and Vertical(HV)fngerprint,the pitch and roll are employed instead of the original fngerprint intensity to extract the horizontal and vertical components of the magnetic feld fngerprint.Derivative dynamic time warping employs the HV fngerprint in its derivative form,which receives higher-level features because of the consideration of fngerprint shape information.Chicken Swarm Optimization(CSO)is used to enhance particle weights,which minimizes position error to tackle the particle impoverishment problem for a fusion navigation system.The results of the experiments suggest that the enhanced algorithm can improve indoor navigation accuracy signifcantly.

Online coherence identification using dynamic time warping for controlled islanding

Controlled islanding is considered to be the last countermeasure to prevent a system-wide blackout in case of cascading failures.It splits the system into self-sustained islands to maintain transient stability at the expense of possible loss of load.Generator coherence identification is critical to controlled islanding scheme as it helps identify the optimal cut-set to maintain the transient stability of the post-islanding systems.This paper presents a novel approach for online generator coherency identification using phasor measurement unit(PMU) data and dynamic time warping(DTW).Results from the coherence identification are used to further cluster non-generator buses using spectral clustering with the objective of minimizing power flow disruptions.The proposed approach is validated and compared to existing methods on the IEEE39-bus system and WECC 179-bus system, through which its advantages are demonstrated.

中国式现代化视域下的测度、演化与模式

文章基于新发展理念并结合中国式现代化的五大特征,构建了经济现代化、创新现代化、共享现代化、生态现代化、治理现代化五个维度的中国式现代化指标体系,依据各省统计年鉴和百度指数数据使用熵权法测算2012-2021年中国31个省份的中国式现代化水平。采用核密度估计、σ收敛模型、β收敛模型和基于动态时间规整的面板数据K中心点聚类算法,分析了中国式现代化的动态演化特征、收敛机制以及发展的模式。研究结果表明:2012-2021年中国式现代化水平逐步提升;东部地区的中国式现代化水平最高,中部地区的增速最高,西部和东北地区存在中国式现代化水平低且增长乏力的问题;全国各省份的中国式现代化水平绝对收敛速度为0.7%,条件收敛速度为3%,不存在σ收敛;聚类分析发现存在发展型和变革型两类中国式现代化模式:发展型中国式现代化以经济现代化为抓手,在共享现代化和生态现代化上表现优异;变革型中国式现代化勇于变革制度寻求突破创新,侧重于治理现代化和创新现代化。

一种改进聚类算法的时间序列异常检测方法

时间序列异常检测被广泛应用于民航领域,对飞机快速存取记录器收集的时间序列数据进行异常检测为识别降低安全裕度的事件提供了有力手段。为了提高时间序列异常检测的准确率,提出一种基于改进聚类算法的时间序列异常检测方法。将K-Medoids聚类算法的欧氏距离度量方法替换为动态时间规整距离度量方法,根据样本点与中心点之间的距离判定异常,研究通过飞机飞行参数超限检测测试时间序列异常检测方法的有效性。实验结果表明,与传统聚类算法相比该方法的异常检测准确率和F1分数更高。聚类算法使用动态时间规整度量距离优化了时间序列相似性度量的精度,可以对形态特点相似的时间序列数据更好地聚类,提高了聚类算法的准确性。

基于PCA-ShapeDTW-QWGRU的分布式光伏集群短期功率预测

针对分布式光伏短期功率预测建立基于主成分分析、改进的动态时间规整算法与量子加权门控循环单元(PCAShapeDTW-QWGRU)的集群功率预测模型。针对集群划分不够精细、光伏电站数据蕴含的信息难以捕捉的问题,提出基于主成分分析结合密度聚类算法(PCA-OPTICS)的集群划分方法;针对目前选取代表电站与集群相似性较低的问题,提出基于改进的动态时间规整算法(ShapeDTW)的代表电站的选取方法,利用ShapeDTW度量相似性距离,选取最小值作为代表电站,并利用基于均方根传播梯度下降法优化的量子加权门控循环单元(RMSprop-QWGRU)模型进行预测;为了解决代表电站与集群功率的变换系数转换差异较大的问题,采用实时变换系数对代表电站进行集群功率值预测计算。实验结果表明,所提方法能有效提升光伏集群功率预测的精度。

基于TCN和迁移学习的混凝土坝变形预测方法

混凝土坝变形测点数据丢失或者新增测点测量时间太短都会导致这部分测点的数据量不足,使得变形预测精度受到影响。为了提高这些小数据量测点的变形预测精度,提出了将时域卷积网络(TCN)与迁移学习相结合的变形预测方法。以数据量充足的测点为源域,以缺少数据的测点为目标域,将在源域上训练好的TCN模型的结构和参数迁移到目标域模型中,固定其中的冻结层参数,利用目标域中的数据对目标域模型可调层参数进行调整。同时,采用动态时间规整选择与目标域数据序列相似度最高的监测数据作为最佳源域数据,提升迁移学习效果。工程实例分析表明:迁移学习后的目标域模型的均方根误差和平均绝对误差与利用足量数据训练的TCN模型的预测误差相比,差异仅分别为1.73%和8.09%,小数据量情况下TCN预测模型的精度得到了提高。

An intelligent automatic correlation method of oilbearing strata based on pattern constraints:An example of accretionary stratigraphy of Shishen 100 block in Shinan Oilfield of Bohai Bay Basin,East China

Aiming at the problem that the data-driven automatic correlation methods which are difficult to adapt to the automatic correlation of oil-bearing strata with large changes in lateral sedimentary facies and strata thickness,an intelligent automatic correlation method of oil-bearing strata based on pattern constraints is formed.We propose to introduce knowledge-driven in automatic correlation of oil-bearing strata,constraining the correlation process by stratigraphic sedimentary patterns and improving the similarity measuring machine and conditional constraint dynamic time warping algorithm to automate the correlation of marker layers and the interfaces of each stratum.The application in Shishen 100 block in the Shinan Oilfield of the Bohai Bay Basin shows that the coincidence rate of the marker layers identified by this method is over 95.00%,and the average coincidence rate of identified oil-bearing strata reaches 90.02% compared to artificial correlation results,which is about 17 percentage points higher than that of the existing automatic correlation methods.The accuracy of the automatic correlation of oil-bearing strata has been effectively improved.

一种数据库查询的多标签电能质量混合扰动识别与分类新方法

该文针对电能质量混合扰动的复杂性及当前分类识别的准确率不够高等问题,提出一种数据库查询的多标签电能质量混合扰动分类与识别方法,该方法能够更加科学准确地识别混合扰动,可为电能质量治理、扰动事件责任追究等提供有力决策辅助。首先,该方法基于可调Q因子小波变换(tunable Q-factor wavelet transform,TQWT)和时变均方根(rootmeansquare,RMS)的特征提取方法有效提取扰动信号基频时域特征量,较好地克服了当前基频幅值特征提取准确率不够高的难点问题;其次,提出频域特征曲线分割新方法,高效地提取扰动信号的高频特征曲线;然后,建立基频幅值特征数据库和高频特征曲线数据库;最后,利用快速动态时间规整(dynamictimewarping,DTW)结合多标签的分类思想进行混合电能质量扰动的多标签分类。仿真实验结果表明,新方法具有如下优势:几乎不受限值范围内基频偏移的影响,抗噪性较强,对单一扰动及包含双重、三重、四重扰动在内的27种扰动具有较高的分类准确率。电网实测扰动数据的分析,进一步验证了该方法的扰动识别有效性。

基于改进DTW算法的永磁同步电机失磁故障模拟与诊断

为实现永磁同步电机(permanent magnet synchronous motor,PMSM)失磁故障模拟与失磁故障程度诊断,提出一种基于改进的动态时间规整(dynamic time warping,DTW)算法的故障波形诊断方法。首先,对永磁同步电机进行机理分析和模拟,采集不同程度失磁故障下的电机转速波形数据集。其次,使用改进DTW算法拟合失磁故障程度最佳路径曲线方程。然后,根据定子电流的频率谐波分量确定电机是否发生失磁故障。最后,使用改进DTW算法对失磁故障波形数据进行故障程度诊断。实验表明,该方法对故障波形数据进行处理后,可以快速准确地计算出故障波形的故障程度,该方法具备较强的稳定性和实时性。

基于遥感时序物候特征的耕地非粮化多模式监测方法

耕地非粮化对粮食生产和农业可持续发展构成潜在威胁,精准监测不同的耕地非粮化类型对制定针对性的农业管理政策至关重要。该研究以河北省石家庄市藁城区为研究区,首先采用最大类间方差算法(OTSU)提取果园和耕地范围,然后利用Google Earth Engine(GEE)云计算平台构建了基于Sentinel-2遥感数据的特征集,包括光谱特征、物候特征和NDVI(normalized difference vegetation index)时序特征。结合面向对象分割和随机森林(radom forest, RF)、时间加权的动态时间规整(time-weighted dynamic time warping, TW-DTW)算法,构建了4种不同的分类模式用于提取粮食作物和露天蔬菜、大棚种植等非粮食作物。通过选择最优模式,提取了研究区2019-2022年间不同非粮化类型的空间分布信息,并探讨了不同模式的优点和局限性。结果表明:1)采用面向对象的机器学习模式进行耕地内作物分类的精度最佳,两个生长季内总体精度分别达到93.23%和90.10%,Kappa系数分别达到0.91和0.88;2)基于时间序列匹配的模式在区分粮食作物和其他地类方面表现出较高的准确性,冬小麦、玉米和大豆的用户精度分别高于95.60%、74.70%、82.70%,制图精度分别高于97.70%、86.40%、93.10%;3)利用面向对象的机器学习模式进行耕地非粮化信息提取,在两个作物生长季的总体精度为87.00%和81.00%。分析耕地非粮化结果发现,藁城区2019-2022年的年际性非粮化面积为2 753.09 hm^(2),其中果园占比最高;而季节性非粮化结果显示,秋粮非粮化面积(3 174.86 hm^(2))明显高于夏粮非粮化面积(1 060.27 hm^(2))。该研究利用Sentinel-2时序遥感数据,为一年两熟区耕地非粮化监测提供一种新的思路,可以为制定差异化农业管理政策提供依据。

基于动态规整与改进变分自编码器的异常电池在线检测方法

针对电池生产成组过程中,传统异常检测方法对混入的容量及压差异常电池检测精度低及生产结束后离线异常检测方法效率低等问题,该文提出一种集合长短期记忆变分自编码器与动态时间规整评价的锂电池异常在线检测方法(VAE-LSTM-DTW),实现了异常电池的在线检测,避免了离线异常检测所造成的时间和能源的浪费。该方法首先将长短期记忆网络(LSTM)引入变分自编码器(VAE)模型,训练电池时序数据重构模型;其次,在电池异常检测的度量标准中引入动态时间规整值(DTW),并基于贝叶斯寻优获得最优检测阈值,对每个单体电池重构数据的动态规整值进行异常辨别。实验结果表明,相较该领域传统异常检测方法,VAE-LSTM-DTW模型性能优越,查准率和F1值都得到了较大的提升,具有较高的有效性和实用性。

基于模式约束的油层单元智能自动对比方法——以渤海湾盆地史南油田史深100区块加积式地层对比为例

针对基于数据驱动的地层自动对比方法难以适应侧向沉积相变快及地层厚度差异大的油层单元自动对比这一问题,建立基于模式约束的油层单元智能自动对比方法。该方法提出在油层单元自动对比中引入知识驱动,采用地层发育模式约束油层单元自动对比过程,并将地层模式约束思想引入构建的相似性度量机及改进的条件约束动态时间规整算法,实现了对标志层及各油层单元界面的自动对比。渤海湾盆地史南油田史深100区块的应用表明:与人工对比结果相比,该方法标志层识别吻合率高于95.00%,油层单元识别平均吻合率达90.02%;与已有自动对比方法相比,油层单元识别平均吻合率提升约17个百分点,有效提高了油层单元自动对比精度。