融合对抗训练及全局指针的实体关系联合抽取

实体关系联合抽取旨在从文本中抽取出实体关系三元组, 是构建知识图谱十分重要的步骤之一. 针对实体关系抽取中存在的信息表达能力不强、泛化能力较差、实体重叠和关系冗余等问题, 提出了一种实体关系联合抽取模型RGPNRE. 使用RoBERTa预训练模型作为编码器, 提高了模型的表达信息能力. 在训练过程中引入了对抗训练, 提升了模型的泛化能力. 使用全局指针, 解决了实体重叠的问题. 使用关系预测, 排除不可能的关系, 减少了冗余的关系. 在基于schema的中文医学信息抽取数据集CMeIE上进行的实体关系抽取实验表明, 模型的F1值比基准模型提升了约2个百分点, 在实体对重叠的情况下, 模型的F1值提升了近10个百分点, 在单一实体重叠情况下, 模型的F1值提升了大约1个百分点, 说明该模型能够更准确地提取实体关系三元组, 从而有效提升知识图谱构建的准确度. 在含有1–5个三元组的对比实验中, 在拥有4个三元组的句子中, 模型的F1值提升了约2个百分点, 而在拥有5个及以上三元组的复杂句子中, F1值提升了约1个百分点, 说明该模型能够较好地处理复杂句子场景.     

预应力混凝土柱的施工技术

偏心受压钢筋混凝土柱,采用预应力技术减小或避免钢筋混凝土柱根部裂缝,通过在钢筋混凝土柱受拉区一侧预先施加一定的力,使钢筋混凝土柱内部受力在新的条件下建立平衡.实践证明:该方法能有效地控制柱截面,解决钢筋混凝土柱偏心受压的根部裂缝问题.     

一种通用的继电保护数据库模型

首先分析了公共信息模型在描述保护信息方面的不足,然后以仿真培训系统为实际应用背景,提出了一种通用的继电保护数据库模型,它能够从保护装置的内部结构和外部特征2个方面完整地表达实际的保护装置。该数据库模型允许用户在保护装置级进行保护配置,支持用户通过自定义新型号的保护装置扩展装置模型,支持装置模型升级后保护配置的自动更新,从而大大提高了保护数据库的维护效率。使用该数据库模型实现的各种仿真培训系统现场运行良好、运行维护方便,证明了该模型的有效性。     

基于优化FBD算法的谐波与闪变综合抑制装置的应用与研究

根据当前电网中存在的实际问题,提出对系统进行谐波与闪变综合抑制的必要性.对传统的FBD(Fryze-Buchholz-Dpenbrock)法进行分析和改进.在此基础上针对三相四线制系统对FBD法进行了优化设计,通过增加一个反馈回路,提高了检测方法的实时性,增强了谐波与闪变抑制装置的动态响应特性.在理论研究与仿真分析的基础上,搭建了基于并联型有源电力滤波器的谐波与闪变综合抑制装置的仿真与实验模型,检测算法采用FBD优化算法,电流跟踪控制电路采用基于滞环比较的空间矢量控制策略.仿真与实验效果验证了检测算法与控制策略的有效性和准确性.     

高压直流输电控制保护系统新技术——WIN-TDC

在新一代直流控制保护系统WIN-TDC中,极控、站控和直流保护等核心部分采用了基于64位精简指令集计算机(RISC)、实时操作系统、64位系统总线的工业控制通用平台--SIMATIC-TDC.SIMATIC-TDC的开发环境采用了单一的编程语言,调试软件集成在开发环境中.人机界面采用了通用工控组态软件--WINCC.利用时分复用(TDM)总线技术实现了测量系统的交叉冗余.WIN-TDC已经在澳大利亚的海底电缆直流工程(BASSLINK)中得到了成功的应用.     

厦门广电集团6信道高清转播车改造方案

介绍了厦门广电集团6信道高清转播车改造背景,从实用性、经济性的角度出发,介绍了详细的车体改造方案、系统方案和视音频方案等.     

Bone Tissue Engineering in Rat Calvarial Defects Using Induced Bone-like Tissue by rhBMPs from Immature Muscular Tissues In Vitro

This study aimed to induce bone-like tissue from immature muscular tissue (IMT) in vitro using commercially available recombinant human bone morphogenetic protein (rhBMP)-2, rhBMP-4, and rhBMP-7, and then implanting this tissue into a calvarial defect in rats to assess healing. IMTs were extracted from 20-day-old Sprague-Dawley (SD) fetal rats, placed on expanded polytetrafluoroethylene (ePTFE) with 10 ng/μL each of rhBMP-2, BMP-4, and BMP-7, and cultured for two weeks. The specimens were implanted into calvarial defects in 3-week-old SD rats for up to three weeks. Relatively strong radiopacity was observed on micro-CT two weeks after culture, and bone-like tissue, comprising osteoblastic cells and osteoids, was partially observed by H&E staining. Calcium, phosphorus, and oxygen were detected in the extracellular matrix using an electron probe micro analyzer, and X-ray diffraction patterns and Fourier transform infrared spectroscopy spectra of the specimen were found to have typical apatite crystal peaks and spectra, respectively. Furthermore, partial strong radiopacity and ossification were confirmed one week after implantation, and a dominant novel bone was observed after two weeks in the defect site. Thus, rhBMP-2, BMP-4, and BMP-7 differentiated IMT into bone-like tissue in vitro, and this induced bone-like tissue has ossification potential and promotes the healing of calvarial defects. Our results suggest that IMT is an effective tissue source for bone tissue engineering.     

Stem Cell Models for Cancer Therapy

Metastatic progression of female breast and colon cancer represents a major cause of mortality in women. Spontaneous/acquired resistance to conventional and targeted chemo-endocrine therapy is associated with the emergence of drug-resistant tumor-initiating cancer stem cell populations. The cancer-initiating premalignant stem cells exhibit activation of select cancer cell signaling pathways and undergo epithelial–mesenchymal transition, leading to the evolution of a metastatic phenotype. The development of reliable cancer stem cell models provides valuable experimental approaches to identify novel testable therapeutic alternatives for therapy-resistant cancer. Drug-resistant stem cell models for molecular subtypes of clinical breast cancer and for genetically predisposed colon cancer are developed by selecting epithelial cells that survive in the presence of cytostatic concentrations of relevant therapeutic agents. These putative stem cells are characterized by the expression status of select cellular and molecular stem cell markers. The stem cell models are utilized as experimental approaches to examine the stem-cell-targeted growth inhibitory efficacy of naturally occurring dietary phytochemicals. The present review provides a systematic discussion on (i) conceptual and experimental aspects relevant to the chemo-endocrine therapy of breast and colon cancer, (ii) molecular/cellular aspects of cancer stem cells and (iii) potential stem-cell-targeting lead compounds as testable alternatives against the progression of therapy-resistant breast and colon cancer.     

Next-Generation Grade and Survival Expression Biomarkers of Human Gliomas Based on Algorithmically Reconstructed Molecular Pathways

In gliomas, expression of certain marker genes is strongly associated with survival and tumor type and often exceeds histological assessments. Using a human interactome model, we algorithmically reconstructed 7494 new-type molecular pathways that are centered each on an individual protein. Each single-gene expression and gene-centric pathway activation was tested as a survival and tumor grade biomarker in gliomas and their diagnostic subgroups (IDH mutant or wild type, IDH mutant with 1p/19q co-deletion, MGMT promoter methylated or unmethylated), including the three major molecular subtypes of glioblastoma (proneural, mesenchymal, classical). We used three datasets from The Cancer Genome Atlas and the Chinese Glioma Genome Atlas, which in total include 527 glioblastoma and 1097 low grade glioma profiles. We identified 2724 such gene and 2418 pathway survival biomarkers out of total 17,717 genes and 7494 pathways analyzed. We then assessed tumor grade and molecular subtype biomarkers and with the threshold of AUC > 0.7 identified 1322/982 gene biomarkers and 472/537 pathway biomarkers. This suggests roughly two times greater efficacy of the reconstructed pathway approach compared to gene biomarkers. Thus, we conclude that activation levels of algorithmically reconstructed gene-centric pathways are a potent class of new-generation diagnostic and prognostic biomarkers for gliomas.