基于活动序列的服务二元行为模式及操作语义

组合E-服务是电子商务开发部署中被广泛关注的问题,行为需求的刻画、建模是服务设计开发的重要步骤,分析和验证服务行为满足用户特定需求则是保证服务可信的关键工作.已有分析服务行为研究中,行为需求都是LTL类逻辑规范或MSC类图式规范,前者基于活动的,后者基于场景的.现在提出一种新颖的基于活动序列的行为规范概念,该概念的粒度介于活动和场景概念之间,目的是便于直接表达某些行为需求,针对活动序列设计了"序列前提、序列因果和序列嵌入"二元行为模式,用模式来刻画基于活动序列的活性和安全性,其结果是行为需求表达更为简洁,将这些模式映射标记传递系统LTS,精确刻画了行为模式的操作语义,为需求验证提供基础,最后给出实例分析.     

单峰映射揉搓序列的特性

为了更好地解释Feigenbaum现象,研究了单峰映射的揉搓序列,利用分析的方法证明了单峰映射揉搓序列极大性的一些性质,进而给出了一个序列作为允许揉搓序列的一个等价条件;利用单峰映射厂和Tpf的揉搓序列之间的关系,给出了重正化算子Tp不动点的揉搓序列的特征．     

基于CAZAC序列的OFDM时频同步算法

针对Schmidl和Minn算法存在符号定时和频率同步精度不高,频率同步运算量大的缺点,提出一种新的基于CAZAC序列的OFDM同步算法.算法利用2个相同的CAZAC序列作为一个训练符号,并对第二个序列用CAZAC序列进行加权,在接收端利用训练符号所包含数据的重复性和CAZAC序列的相关性,进行符号定时同步;然后利用CAZAC序列具有的时域移位特性进行整数倍频偏估计.算法在时域完成,因运算量较小.仿真结果表明,新算法克服了Schmidl和Minn算法的缺点,在AWGN和多径信道下,均能精确同步并进行较大范围的频偏估计.特别是在信噪比较低时采用该算法能实现更准确的同步.     

一种结合猫映射与L og ist ic 映射的语音加密算法

将猫映射(cat map)与Logistic映射相结合，构造了一种语音加密算法．该算法首先将语音数据堆叠成二维，然后利用二维猫映射将数据的位置置乱，最后利用一维Logistic映射构造替换表，对数据进行扩散．密码分析表明，该算法具有较高的安全性，能够抵抗统计攻击、差分攻击和已知密文攻击．与传统的DES算法相比，该算法加密速度更快，适用于实时语音加密．     

基于“双芯”51内核的学习机设计

嵌入式系统在教育领域的应用产生了各种类型的学习机,重点论述了一种基于M CS-51内核的“双芯”的嵌入式学习机系统,满足了职业学校学生录入练习的需求。通过对这个系统的分析和论述提出一种在51内核上运行复杂系统的方法。     

基于掩模金字塔的高精度全局运动估计算法

在视频序列的全局运动估计中,前景运动对象的存在常常会大幅度地降低估计的精度,为此提出一种对前景对象自适应的高精度全局运动估计算法.该算法以像素块为单位,利用块内外点的比重判定前景区域,同时引入马尔可夫聚类方法进行后处理,有效地提高了运动对象的定位精度;通过对目标函数引入权重系数增强对残差的鲁棒性,以进一步提高算法的估计精度.此外,基于像素掩模的3层金字塔构建序列图像,并将改进的梯度方法引入到优化过程中,提高了算法的实时性.对不同运动类型的标准视频序列的实验结果表明,该算法有效地提高了全局运动估计的精度和速度.     

异构软件体系结构的可靠性模型

现有基于软件体系结构的可靠性模型只适用于符合马尔可夫过程的软件构件转移关系,而当构件间存在由确定性和随机性转移关系组成的异构结构时,仍按照马尔可夫过程对软件体系结构建模,会导致可靠性评估结果不准确。为了解决现有模型不适用于异构结构的问题,在分析构件转移关系基础上,提出了一个新的软件可靠性模型。基于软件UML顺序图,提出将异构软件结构转换为马尔可夫链的算法。通过对一个实例的可靠性评估,验证了新模型能够比现有模型获得更准确的评估结果。     

Genomic Insights of First ermB-Positive ST338-SCCmecVT/CC59 Taiwan Clone of Community-Associated Methicillin-Resistant Staphylococcus aureus in Poland

We report the first Polish representative of community-associated methicillin-resistant Staphylococcus aureus (CA-MRSA), lukS/F-PV-positive, encoding the ermB gene, as a genetic determinant of constitutive resistance to macrolides, lincosamides, and streptogramin B antibiotics, cMLS-B. This is the first detection of the CA-MRSA strain responsible for nosocomial infection in the Warsaw Clinical Hospital. Resistance to β-lactams associates with a composite genetic element, SCCmec cassette type V_T (5C2&5). We assigned the strain to sequence type ST338 (single-locus variant of ST59), clonal complex CC59, spa-type t437, and agr-type I. Genomic-based comparison was designated SO574/12 as an international Taiwan clone, which has been so far described mainly in the Asia-Pacific region. The ermB gene locates on the chromosome within the 14,690 bp mobile element structure, i.e., the MES_PM1-like structure, which also encodes aminoglycoside- and streptothricin-resistance genes. The MES_PM1-like structure is a composite transposon containing Tn551, flanked by direct repeats of IS1216V insertion sequences, which probably originates from Enterococcus. The ermB is preceded by the 273 bp regulatory region that contains the regulatory 84 bp ermBL ORF, encoding the 27 amino acid leader peptides. The latest research suggests that a new leader peptide, ermBL2, also exists in the ermB regulatory region. Therefore, the detailed function of ermBL2 requires further investigations.     

改进Informer网络的风电功率短期预测

准确预测风电功率对于提高电力系统的效率和安全性具有重要意义, 而风能的间歇性和随机性特点导致风电功率难以准确预测. 因此, 提出一种改进Informer的风电功率预测模型PCI-Informer (PATCH-CNN-IRFFN-Informer). 将序列数据划分为子序列级补丁, 并进行特征提取和整合, 提高模型对序列数据的处理能力和效果; 采用多尺度因果卷积自注意力机制, 实现多尺度局部特征融合, 提高模型对局部信息的理解和建模能力; 引入反向残差前馈网络 (IRFFN), 增强模型对局部结构信息的提取和保留能力. 某风电场数据实验结果表明, 与主流预测模型相比, PCI-Informer模型在不同预测步长下均取得了更好的预测效果, 在MAE指标上相比Informer模型平均降低了11.1%, 有效提高了短期风电功率的预测精度.     

Genomic and Immunological Characterization of Hypermucoviscous Carbapenem-Resistant Klebsiella pneumoniae ST25 Isolates from Northwest Argentina

In recent years, an increase in the prevalence hypermucoviscous carbapenem-resistant Klebsiella pneumoniae with sequence type 25 (ST25) was detected in hospitals of Tucuman (Northwest Argentina). In this work, the virulence and the innate immune response to two K. pneumoniae ST25 strains (LABACER 01 and LABACER 27) were evaluated in a murine model after a respiratory challenge. In addition, comparative genomics was performed with K. pneumoniae LABACER01 and LABACER27 to analyze genes associated with virulence. Both LABACER01 and LABACER27 were detected in the lungs of infected mice two days after the nasal challenge, with LABACER01 counts significantly higher than those of LABACER27. Only LABACER01 was detected in hemocultures. Lactate dehydrogenase (LDH) and albumin levels in bronchoalveolar lavage (BAL) samples were significantly higher in mice challenged with LABACER01 than in LABACER27-infected animals, indicating greater lung tissue damage. Both strains increased the levels of neutrophils, macrophages, TNF-α, IL-1β, IL-6, KC, MCP-1, IFN-γ, and IL-17 in the respiratory tract and blood, with the effect of LABACER01 more marked than that of LABACER27. In contrast, LABACER27 induced higher levels of IL-10 in the respiratory tract than LABACER01. Genomic analysis revealed that K. pneumoniae LABACER01 and LABACER27 possess virulence factors found in other strains that have been shown to be hypervirulent, including genes required for enterobactin (entABCDEF) and salmochelin (iroDE) biosynthesis. In both strains, the genes of toxin–antitoxin systems, as well as regulators of the expression of virulence factors and adhesion genes were also detected. Studies on the genetic potential of multiresistant K. pneumoniae strains as well as their cellular and molecular interactions with the host are of fundamental importance to assess the association of certain virulence factors with the intensity of the inflammatory response. In this sense, this work explored the virulence profile based on genomic and in vivo studies of hypermucoviscous carbapenem-resistant K. pneumoniae ST25 strains, expanding the knowledge of the biology of the emerging ST25 clone in Argentina.