Improved Relative-transformation Principal Component Analysis Based on Mahalanobis Distance and Its Application for Fault Detection

主要部件分析(PCA ) 广泛地在过程工业被使用了，它能维持最大的差错察觉率。尽管许多问题在 PCA 被处理了，一些必要问题仍然保持未解决。这研究以下列方法为差错察觉性能改进 PCA。第一，一个相对转变计划基于 Mahalanobis 距离(MD ) 被介绍消除数据的尺寸的效果而不是无尺寸的标准化，并且改进精确性和差错察觉的即时性能。理论推导证明那相对转变能直接基于 MD 消除尺寸的效果并且在结果显示出的相对空间，分析和模拟给 PCA 的合理解释它的优势和有效性。第二，一个改进摆平的预言错误(SPE ) 统计数值被给改进标准化 PCA 的差错察觉表演，它能使标准化基于 PCA 的差错察觉方法成为对实际工业过程合适的更多。最后，二个改进方法被联合更有效地检测差错。建议方法被使用在热连续滚动过程检测 looper 系统的单个差错和多差错，模拟结果以易感知，精确性和差错察觉的即时性能为差错察觉性能表明这些改进的有效性。

Euclidean Distance Transform on the Sea Based on Cellular Automata Modeling

To explore the problem of distance transformations while obstacles existing,this paper presents an obstacle-avoiding Euclidean distance transform method based on cellular automata.This research took the South China Sea and its adjacent sea areas as an example,imported the data of land-sea distribution and target points,took the length of the shortest obstacle-avoiding path from current cell to the target cells as the state of a cellular,designed the state transform rule of each cellular that considering a distance operator,then simulated the propagation of obstacle-avoiding distance,and got the result raster of obstacle-avoiding distance transform.After analyzing the effect and precision of obstacle avoiding,we reached the following conclusions:first,the presented method can visually and dynamically show the process of obstacle-avoiding distance transform,and automatically calculate the shortest distance bypass the land;second,the method has auto-update mechanism and each cellular can rectify distance value according to its neighbor cellular during the simulation process;at last,it provides an approximate solution for exact obstacle-avoiding Euclidean distance transform and the proportional error is less than 1.96%.The proposed method can apply to the fields of shipping routes design,maritime search and rescue,etc.

Transmission Lines Distance Protection Using Differential Equation Algorithm and Hilbert-Huang Transform

This paper proposed the scheme of transmission lines distance protection based on differential equation algorithms (DEA) and Hilbert-Huang transform (HHT). The measured impedance based on EDA is affected by various factors, such as the distributed capacitance, the transient response characteristics of current transformer and voltage transformer, etc. In order to overcome this problem, the proposed scheme applies HHT to improve the apparent impedance estimated by DEA. Empirical mode decomposition (EMD) is used to decompose the data set from DEA into the intrinsic mode functions (IMF) and the residue. This residue has monotonic trend and is used to evaluate the impedance of faulty line. Simulation results show that the proposed scheme improves significantly the accuracy of the estimated impedance.

Hardware Euclidean Distance Transform Implementation

Based on an efficient algorithm of Euclidean distance transform for binary images, a circuit of O(N2) size is proposed. With in-place calculation, both the intermediate data storing and the result output use the same memory with the input data. This reduces the amount of memory largely. By replacing multipliers with counters, comparators, and adders, the circuit size is further reduced and its calculation speed is improved also.

Distance Transform Based Enhancement for Linear Interpolated Images

An approach of distane map based imageenhancement (DMIE) is proposed. It is applied toconventional interpolations to get sharp images. Edgedetection is performed after images are interpolatedby linear interpolations. To meet the two conditionsset for DMIE, i. e., no abrupt changes and no over-boosting, different boosting rate should be used inadjusting pixel intensities. When the boosting rate isdetermined by using the distance from enhancedpixels to nearest edges, edge-oriented imageenhancement is obtained. By using Erosion technique,the range for pixel intensity adiustment is set.Over-enhancement is avoided by limiting the pixel iutensities in enhancement within the range. A unifled linear-time algoritiml for disance transform is adopted to deal with the calculation of Euelidean distance of the images.Its computation complexity is 0(N).After the preparation,i.e.,distance transforming and erosion,the images get more and more sharpened while no over.boosting.Occurs by repeating the enhancement procedure ,The simplicity of the enhancement operation makes DMIE suitable for enhancement rate adjusting

基于Transformer复杂运动辨识的机动星凸形扩展目标跟踪方法

针对复杂的机动扩展目标跟踪问题,利用Transformer网络设计了一种有效的星凸不规则形状机动扩展目标跟踪方法。首先,该文研究利用alpha-shape算法建立了星凸形状的变化模型,实现了静态场景下的星凸形扩展目标的形状估计。然后,通过对目标状态转移矩阵进行重新设计,结合Transformer网络对机动扩展目标运动状态转移矩阵进行实时估计,实现了对复杂机动目标运动过程的精准跟踪。进一步地,将估计得到的形状轮廓与运动状态进行融合,最终实现了对星凸形机动扩展目标的实时跟踪。最后,通过构造复杂的机动扩展目标跟踪场景,利用多重性能指标测试算法对形状和运动状态的综合估计性能,验证了算法的有效性。

A SKELETONIZATION ALGORITHM BASED ON EUCLIDEAN DISTANCE MAPS AND MORPHOLOGICAL OPERATORS

In this letter a new skeletonization algorithm is proposed. It combines techniques of fast construction of Euclidean Distance Maps(EDMs), ridge extraction, Hit-or-Miss Transformation(HMT) of structuring elements and the set operators. It first produces the EDM image with no more than 4 passes through an image of any kinds, and then the ridge image is extracted by applying a turn-on scheme and performing a rain-fall elimination to accelerate the processing. The one-pixel wide skeleton is finally acquired by carrying out the HMTs of two structure elements and the SUBTRACT and OR operations. Experimental results obtained by practical applications are also presented.

基于Hausdorff Distance算法的配电网故障区段定位方法

为解决配电网小电流接地系统发生单相接地故障时故障区段定位困难的问题,结合故障线路中故障点两侧零序电流方向相反的特点,提出一种基于Hausdorff Distance算法的配电网故障区段定位方法。该方法将零序电流作为故障特征量,首先,对零序电流进行滤波,并通过小波包变换提取故障馈线各检测节点的零序电流小波近似序列;然后,借助Hausdorff Distance算法分析得到各检测节点之间的零序电流小波近似序列偏差矩阵;最后,综合定义一种偏差度表征各区段两侧的差异性,通过比较偏差度大小确定故障区段。仿真结果表明:该方法在不同的故障条件下均能实现正确定位,同时还适用于复杂的配网结构中,便于配电网运行人员迅速对故障线路进行检修和维护,保障电力系统安全可靠运行。

A novel shapelet transformation method for classification of multivariate time series with dynamic discriminative subsequence and application in anode current signals

Classification of multi-dimension time series(MTS) plays an important role in knowledge discovery of time series. Many methods for MTS classification have been presented. However, most of these methods did not consider the kind of MTS whose discriminative subsequence was not restricted to one dimension and dynamic. In order to solve the above problem, a method to extract new features with extended shapelet transformation is proposed in this study. First, key features is extracted to replace k shapelets to calculate distance, which are extracted from candidate shapelets with one class for all dimensions. Second, feature of similarity numbers as a new feature is proposed to enhance the reliability of classification. Third, because of the time-consuming searching and clustering of shapelets, distance matrix is used to reduce the computing complexity. Experiments are carried out on public dataset and the results illustrate the effectiveness of the proposed method. Moreover, anode current signals(ACS) in the aluminum reduction cell are the aforementioned MTS, and the proposed method is successfully applied to the classification of ACS.

CTUNING:A REUSE DISTANCE BASED CACHE PERFORMANCE TUNING TOOL

Cache performance tuning tools are conducive to develop program with good locality and fully use cache to decrease the influence caused by speed gap between processor and memory. This paper introduces the design and implementation of a cache performance tuning tool named CTuning, which employs a source level instrumentation method to gather program data access information, and uses a limited reuse distance model to analyze cache behavior. Experiments on 183.equake improve average performance more than 6% and show that CTuning is proficient not only in locating cache performance bottlenecks to guide manual code transformation, but also in analyzing cache behavior relationship among variables, thus to direct manual data reorganization.

Handwritten Character Recognition Using Multiresolution Technique and Euclidean Distance Metric

In the present paper, the problem of handwritten character recognition has been tackled with multiresolution technique using discrete wavelet transform (DWT) and Euclidean distance metric (EDM). The technique has been tested and found to be more accurate and faster. Characters is classified into 26 pattern classes based on appropriate properties. Features of the handwritten character images are extracted by DWT used with appropriate level of multiresolution technique, and then each pattern class is characterized by a mean vector. Distances from input pattern vector to all the mean vectors are computed by EDM. Minimum distance determines the class membership of input pattern vector. The proposed method provides good recognition accuracy of 90% for handwritten characters even with fewer samples.

基于Transformer与距离图谱的泛癌细胞核图像分割

肿瘤细胞的密度、核质比和平均尺寸等指标对癌症的分级和预后有重要的意义.在计算病理学中,细胞核分割是肿瘤微环境分析的基础.通过对分割结果进行统计分析,对新的肿瘤标记物的探索有重大的意义.病理图像背景下的细胞核形态不规则,细胞核染色不均匀,且细胞核边缘之间存在黏连的问题,而现有的深度学习算法在细胞核主体分割正确的情况下,边缘的分割错误不会对总体的损失造成太大的影响,黏连的细胞核很容易被当作同一个分割目标.为了解决细胞核重叠问题,本文提出一种基于Transformer与距离图谱的分割模型,将Transformer中的核心多头自注意力机制与距离图谱引导算法相结合,重视细胞核内部,弱化细胞核边界,提升对图像局部和全局上下文信息的学习能力.本文方法在两个公开数据集上的平均Dice系数为0.7979、精度为0.7561、AJI系数为0.6672、Hausdorff距离为10.11.实验结果表明,相较其他分割算法,本文方法的性能更好,能够有效提高细胞核的分割精度,同时较好地解决了细胞核之间的黏连问题.

Deep neural network based on multi-level wavelet and attention for structured illumination microscopy

Structured illumination microscopy(SIM)is a popular and powerful super-resolution(SR)technique in biomedical research.However,the conventional reconstruction algorithm for SIM heavily relies on the accurate prior knowledge of illumination patterns and signal-to-noise ratio(SNR)of raw images.To obtain high-quality SR images,several raw images need to be captured under high fluorescence level,which further restricts SIM’s temporal resolution and its applications.Deep learning(DL)is a data-driven technology that has been used to expand the limits of optical microscopy.In this study,we propose a deep neural network based on multi-level wavelet and attention mechanism(MWAM)for SIM.Our results show that the MWAM network can extract high-frequency information contained in SIM raw images and accurately integrate it into the output image,resulting in superior SR images compared to those generated using wide-field images as input data.We also demonstrate that the number of SIM raw images can be reduced to three,with one image in each illumination orientation,to achieve the optimal tradeoff between temporal and spatial resolution.Furthermore,our MWAM network exhibits superior reconstruction ability on low-SNR images compared to conventional SIM algorithms.We have also analyzed the adaptability of this network on other biological samples and successfully applied the pretrained model to other SIM systems.

融合Transformer和MSCNN双分支架构的工控网络入侵检测研究

针对现有的工控网络入侵检测方法中存在对工控流量的多空间特征和长距离时序特征的提取能力不足等问题,提出了一种融合Transformer和MSCNN双分支架构的工控网络入侵检测模型.该模型利用多尺度卷积(MSCNN)中多个不同大小卷积核,对工控流量中多个空间特征进行抽取,扩大了对工控流量特征范围的学习.同时引入Transformer增强了模型对工控流量中长距离时序特征的提取能力,进一步提高了模型的性能.通过UNSW-NB15和NSL-KDD数据集进行了实验,结果表明:该模型与其他方法相比能够提取更加全面有效的特征,具有很好的检测性能和泛化能力.

Lecture Notes of Mobuis Transformation in Hyperbolic Plane

In this paper, I have provided a brief introduction on M?bius transformation and explored some basic properties of this kind of transformation. For instance, M?bius transformation is classified according to the invariant points. Moreover, we can see that M?bius transformation is hyperbolic isometries that form a group action PSL (2, R) on the upper half plane model.

领域特征融合Transformer的环焊缝缺陷识别方法

环焊缝缺陷的类型识别对于长输管线管道焊缝质量评价以及管道服役寿命评估具有重要意义。由于不同类型缺陷的射线检测图像具有特征差异小、对比度低等特征,现有缺陷类型识别方法所提取的特征表征能力不足,其准确性及可信性难以满足行业需求。为此,提出了一种缺陷领域特征提取与Transformer融合的焊缝缺陷类型识别方法。基于环焊缝缺陷评定技术人员的知识及缺陷领域特征,从缺陷几何特征、缺陷位置特征以及缺陷背景区域特征3个方面定义了14个特征,用于实现不同类型缺陷的特征表征;以Transformer网络为基础,融合上述14个特征提出深度可调节的缺陷类型识别模型与方法。以企业实际环焊缝缺陷数据对模型进行验证。结果表明,与ResNet50网络相比,所提模型对于未熔合、未焊透缺陷,分类精度分别提高18.2%和14.3%;对于咬边和内凹2种形状缺陷,分类精度达到90%以上。证明所提方法可以有效提高缺陷类型识别准确率,可将其扩展到射线检测铸造缺陷、TOFD检测焊缝缺陷识别领域。

基于SF-Transformer的智能教育平台短期电力负荷预测研究

建设智能教育平台是推动教育智能化的一个重要过程,但智能教育平台依赖的人工智能模型在训练过程中会消耗大量电力,因此,开展短期电力负荷预测对建设智能教育平台具有重要意义.针对在考虑多个属性开展短期电力负荷预测时,由于部分属性与电力负荷数据的相关性不强并且Transformer无法捕捉电力负荷数据的时间相关性,而导致电力负荷预测不够准确的问题,基于SR(Székely and Rizzo)距离相关系数、融合时间定位编码和Transformer,提出了一种短期电力负荷预测模型SF-Transformer.SF-Transformer通过SR距离相关系数对影响电力负荷数据的属性进行筛选,选择与电力负荷数据之间SR距离相关系数较大的属性.SF-Transformer采用一种全局时间编码与局部位置编码相结合的融合时间定位编码,有助于模型全面获取电力负荷数据的时间定位信息.在数据集上开展了实验,实验结果表明SF-Transformer与其他模型相比,在两种时长上进行电力负荷预测具有更低的均方根误差和平均绝对误差.

利用Transformer的组合聚类算法在蛋白质数据分析中的应用

该研究将Transformer模型适配于蛋白质特征降维场景,通过其特有的自注意力机制,赋予模型对长程依赖关系的较好建模性能,同时,多头注意力设计使得模型能够从不同角度捕获特征间的相互作用,进一步提升降维结果的表达力和鲁棒性。文章提出了一种新型的GRKM组合聚类算法,在原始K-means算法中引入了灰狼优化算法(Grey Wolf Optimization Algorithm)确定聚类的K值,以随机游走算法(Random Walk)确定初始聚类中心,以马氏距离(Markov Distance)来衡量样本间的相似性。研究中,对5种具有代表性的蛋白质数据集进行了实验验证,得到了改进后算法在轮廓系数以及DB指数等方面相较于改进前都有较大提升的结论。最终的结果分析选取APP蛋白质数据,将蛋白质聚为8类,探讨了各类别的生物功能,在解释性方面也取得了较为明显的效果。所提算法为深入理解蛋白质功能、发现潜在生物标志物以及指导药物设计等实际应用提供了参考工具。

基于Transformer-Bi-LSTM模型的武器装备剩余寿命预测方法

武器装备担负保卫国土安全的重要使命,其保持稳定运行状态具有重大国防、政治意义;因其装备运行状态不便中断、故障定位过程复杂,使得传统维修方式效率较低;装备使用数据具有连续性、长期性、不平稳性,甚至一些深度学习模型无法处理其中的退化状态历史依赖与关联问题;通过构建元器件层级的剩余寿命预测架构,对特征工程、退化指标构建以及Transformer-Bi-LSTM模型开展研究,采用距离编码技术,实现针对深度学习模型的技术创新,优化模型预测效果;基于某型武器装备主要器件正常试样数据,进行本方法分析验证,在器件已运行时间达到90%设计试验寿命长度时能够进行有效且准确的剩余寿命预测,所提方法满足武器装备器件寿命预警及更换提醒、保障装备战备完好性的应用需求。

Simultaneous measurement of distance and speed using frequency-modulated continuous-wave self-mixing interferometry and all-phase fast Fourier transform

In this work,we propose a method using frequency-modulated continuous-wave(FMCW)self-mixing interferometry(SMI)and all-phase fast Fourier transform(APFFT)for simultaneous measurement of speed and distance.APFFT offers superior accuracy in frequency determination by mitigating issues like the fence effect and spectrum leakage,contributing to the high-accuracy measurement for speed and distance.Both simulations and experiments have demonstrated relative errors at the levels of 10^(−4) and 10^(−3) for distance and speed measurements,respectively.Furthermore,factors impacting measurement performance have been discussed.The proposed method provides a high-performance and cost-effective solution for distance and speed measurements,applicable across scientific research and various industrial domains.