Image sequence-based risk behavior detection of power operation inspection personnel

A novel image sequence-based risk behavior detection method to achieve high-precision risk behavior detection for power maintenance personnel is proposed in this paper. In this method, the original image sequence data is first separated from the foreground and background. Then, the free anchor frame detection method is used in the foreground image to detect the personnel and correct their direction. Finally, human posture nodes are extracted from each frame of the image sequence, which are then used to identify the abnormal behavior of the human. Simulation experiment results demonstrate that the proposed algorithm has significant advantages in terms of the accuracy of human posture node detection and risk behavior identification.     

Cloning,sequencing and structural analysis of membrane-bound polyphenol oxidase from Granny Smith apples (Malus × domestica Borkh)

The gene sequence coding for the membrane-bound polyphenol oxidase (mPPO) with a length of 1761 bp was cloned by PCR method and shown to contain one highly conserved sequence encoding a di-copper-binding region. The predicted three-dimensional structure of mPPO indicated that the active site was located near two copper ions and composed of a typical bundle of four α-helices. Each of the two catalytic copper ions was coordinated with three histidine residues in the hydrophobic pocket, yielding His 180, His 201, His 210, His 332, His 336 and His 366. Docking studies showed that 4-methylcatechol and chlorogenic acid have different binding models due to different ligand sizes and binding sites in the active centre, and it was found that the smaller compound exhibited a higher affinity for mPPO. Molecular dynamic simulation results indicated that Phe 353 is important in controlling enzymatic activity through influencing substrate coordination in the active site.     

Compilation of Modern Technologies To Map Genome-Wide Cytosine Modifications in DNA

Over the past few decades, various DNA modification detection methods have been developed; many of the high-resolution methods are based on bisulfite treatment, which leads to DNA degradation, to a degree. Thus, novel bisulfite-free approaches have been developed in recent years and shown to be useful for epigenome analysis in otherwise difficult-to-handle, but important, DNA samples, such as hmC-seal and hmC-CATCH. Herein, an overview of advances in the development of epigenome sequencing methods for these important DNA modifications is provided.     

Effect of ZnO/K2O ratio on the crystallization sequence and microstructure of lithium disilicate glass-ceramics

The effect of ZnO/K₂O (Z/K) ratio on the crystallization sequence and microstructure of lithium disilicate (Li₂Si₂O₅: LS2) glass-ceramics was carefully investigated for the SiO₂-Li₂O-K₂O-ZnO-P₂O₅ system. The Z/K ratios of precursor glasses were varied from 0 to 3.5 while the nucleating agent of P₂O₅ and glass modifiers of ZnO plus K₂O were fixed to have 1.5 and 4.5 mol% relative to LS2, respectively. For the samples prepared by two-stage heat treatments of 500 °C for 1 h and 800 °C for 2 h in air, the LS2 nucleation rate was increased with increasing the Z/K ratio due to the variation in crystallization sequence from type II (Li₂SiO₃: LS) to type I (LS + LS2) in addition to an amorphous phase separation in base glass. Consequently, with increasing the Z/K ratio, the LS2 crystalline phase within the glass matrix continuously changed from larger acicular ones to smaller equiaxed ones.     

基于深度学习的文本自动摘要方案

针对自然语言处理（NLP）生成式自动摘要领域的语义理解不充分、摘要语句不通顺和摘要准确度不够高的问题，提出了一种新的生成式自动摘要解决方案，包括一种改进的词向量生成技术和一个生成式自动摘要模型。改进的词向量生成技术以Skip-Gram方法生成的词向量为基础，结合摘要的特点，引入词性、词频和逆文本频率三个词特征，有效地提高了词语的理解；而提出的Bi-MulRnn+生成式自动摘要模型以序列映射（seq2seq）与自编码器结构为基础，引入注意力机制、门控循环单元（GRU）结构、双向循环神经网络（BiRnn）、多层循环神经网络（MultiRnn）和集束搜索，提高了生成式摘要准确性与语句流畅度。基于大规模中文短文本摘要（LCSTS）数据集的实验结果表明，该方案能够有效地解决短文本生成式摘要问题，并在Rouge标准评价体系中表现良好，提高了摘要准确性与语句流畅度。     

考虑决策者时序偏好的时域证据融合方法

针对时域不确定信息的融合难题，为充分体现时域信息融合的动态性特点和时间因素对融合的影响，在证据理论的基础上，提出一种考虑决策者时序偏好的时域证据融合方法。首先将决策者对时序的偏好融入时域证据融合，通过分析时域证据序列的特点，在定义时序记忆因子的基础上，对决策者的时序偏好进行度量；然后通过构建优化模型求解时序权重，再结合证据信任度的概念，对证据源进行修正；最后利用Dempster组合规则对修正后的证据进行融合。数值算例表明，与没有考虑时间因素的融合方法相比，考虑决策者时序偏好的证据融合方法可以有效处理时域信息序列中的冲突信息，得到合理的融合结果；同时，所提方法充分考虑了时域证据序列的信任度和决策者的主观偏好，可以反映决策者主观因素对时域证据融合的影响，较好地体现了时域证据融合的动态性特点。     

Combination of nuclear magnetic resonance spectroscopy and nonlinear methods to analyze the copolymerization of phosphonic acid derivatives

We demonstrate in this study that the combination of modern inline monitoring methods [here: inline nuclear magnetic resonance (NMR)] with simulations gains more exact and profound kinetic results than previously used methods like linearization without that combination. The ¹H-NMR spectroscopic data (more than 100 data points) are used to construct the copolymerization diagram. The reactivity ratios are obtained applying the van Herks nonlinear least square method. The examination of the radical copolymerization of 2-hydroxyethyl methacrylate (HEMA) with (2-{[2-(ethoxycarbonyl)prop-2-en-1-yl]oxy}ethyl) phosphonic acid (ECPPA) as important adhesive monomer used in dentistry yields reactivity ratios of r_HEMA = 1.83; r_ECPPA = 0.42. The copolymerization diagram reflects nonideal, non-azeotropic copolymerization. The sequence distribution of the obtained by Monte Carlo simulation indicates the generation of statistical copolymers. As an important finding, it is demonstrated that the repeating units responsible for etching and adhesion are arranged over the whole polymer chain, which is necessary to achieve proper functionality. © 2019 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2019 , 136, 48256.     

The cumulative power of incremental innovation and the role of project sequence management

In innovation and project management studies incremental development projects are perceived as theoreticlly and organisationally uninteresting. By means of a longitudinal study of product improvement projects at an automobile firm, this paper challenges such views and shows how the cumulative impact of the studied sequence resulted in a competitive repositioning of the company's product portfolio during a financially difficult period. Project managers achieved this by transcending the separation between exploration and exploitation projects; they not only adhered to time, cost and quality goals but also tried out new ways of testing and experimenting with controversial technical ideas. The paper analyzes the intensive inter project learning that generated these ambidextrous capabilities and emphasizes that practices at the project-level need to be buttressed by expanded management learning and capability development also at the sequence level.     

改进卷积输入的端到端普通话语音识别

主流神经网络训练的交叉熵准则是对声学数据的每个帧进行分类优化，而连续语音识别是以序列级转录准确性为性能度量。针对这个不同，构建基于序列级转录的端到端语音识别系统。针对低资源语料条件下系统性能不佳的问题，其中模型使用卷积神经网络对输入特征进行处理，选取最佳的网络结构，在时域和频域进行二维卷积，从而改善输入空间中因不同环境和说话人产生的小扰动影响。同时神经网络使用批量归一化技术来减少泛化误差，加速训练。基于大型的语言模型，优化解码过程中的超参数，提高模型建模效果。实验结果表明系统性能提升约24%，优于主流语音识别系统。