基于DAE-BP神经网络的股票预测研究

股票指标数据种类多、维度高，且指标之间存在多重共线性。为了降低数据的维度、消除指标间的多重共线性和预测股票价格，首先构建了基于受限布尔兹曼机的深度自编码器，实现了高维数据向低维空间的压缩编码。然后基于BP神经网络建立了低维编码序列与股票价格之间的回归模型。实验结果表明，深度自编码器提取特征的能力优于主成分分析法和因子分析法；相比较使用降维前的数据，使用编码后的数据用预测股票价格，模型可以减少计算开销，并且获得更高的预测精度。     

医院级CT辐射剂量及诊断剂量水平研究

探讨医院级各部位CT扫描辐射剂量及成人CT检查诊断剂量水平。采集2018年2月—2019年1月本院在GE Light Speed 16 CT设备上进行诊断性CT检查的全部成年人共15 440例的CT检查数据,按照检查部位进行分类,统计各部位的容积剂量指数(CTDI_vol)、剂量长度乘积(DLP)。使用SPSS 19统计软件分析CTDI_vol、DLP的四分位数并计算有效剂量(ED)。结果表明,各部位CTDI_vol的诊断剂量参考水平(DRL)值为:颅脑87.75 mGy、肺部8.09 mGy、上下腹15.82 mGy、上下腹盆腔15.79 mGy;各部位DLP的诊断剂量参考水平(DRL)值为:颅脑1 053.03 mGy·cm、肺部245.19 mGy·cm、上下腹269.96 mGy·cm、上下腹盆腔835.20 mGy·cm。研究中发现肺部CTDI_vol、DLP的DRL值(分别为8.09 mGy、245.19 mGy·cm)低于我国其他地区和其他国家的DRL值;60岁以下患者肺部CT检查CTDI_vol、DLP的DRL值低于8.09 mGy、245.19 mGy·cm,60岁以上患者肺部CT检查CTDI_vol、DLP的DRL值高于8.09 mGy、245.19 mGy·cm。通过研究可以确定医院级各部位CT检查辐射剂量及成人CT检查诊断剂量水平,能为优化CT检查扫描技术和监测辐射剂量提供基础。     

激光立体成形TC4钛合金的热力场演化

激光立体成形能够实现高性能复杂零件的精确成形，但成形过程中的高温度梯度极易导致残余应力和变形，最终影响构件的几何精度和力学性能。通过搭建基板单边夹持原位热变形测量系统，探究了多道多层TC4钛合金成形件在激光立体成形过程中基板温度、变形的演化规律，分析了不同扫描策略对构件热应力场的影响。研究结果表明，构件的变形发展受其热历史驱使，扫描策略对成形件的热变形具有显著影响。     

一种简易的磨头喷水装置应用设计

利用简单的材料,结合现场条件,设计制造了一套简易可行的磨头喷水装置,不但能有效降低磨内水泥的温度,更节约了公司的成本以及后期维护更换零配件的费用.     

Biphenyl liquid crystal epoxy containing flexible chain: Synthesis and thermal properties

A biphenyl type liquid crystal epoxy (LCE) monomer 4,4′-di(2,3-epoxyhexyloxy)biphenyl (LCBP4) containing flexible chain was synthesized and the curing behavior was investigated using 4,4′-diaminodiphenylmethane (DDM) as the curing agent. The effect of curing condition on the formation of the liquid crystalline phase was examined. The cured samples show good mechanical properties and thermal stabilities. Moreover, the relationship between thermal conductivity and structure of liquid crystalline domain was also discussed. The samples show high thermal conductivity up to 0.28–0.31 W/(m*K), which is 1.5 times as high as that of conventional epoxy systems. In addition, thermal conductive filler, Al₂O₃, was introduced into LCBP4/DDM to obtain higher thermal conductive composites. When the content of Al₂O₃ was 80 wt%, the thermal conductivity of the composite reached to 1.86 W/(m*K), while that of diglycidyl ether of bisphenol A (Bis-A) epoxy resin/DDM/Al₂O₃ was 1.15 W/(m*K). Compared with Bis-A epoxy resin, the formation of liquid crystal domains in the cured LCE resin enhanced the thermal conductivity synergistically with the presence of Al₂O₃. Furthermore, the introduction of Al₂O₃ also slightly increased the thermal stabilities of the cured LCE.     

Multi-granularities capture interaction information for text matching accuracy enhancement

Modeling and Matching texts is a critical issue in natural language processing (NLP) tasks. In order to improve the accuracy of text matching, multi-granularities capture matching features (MG-CMF) model was proposed. The proposed model used convolution operations to construct the representation of text under multiple granularities, used max-pooling operations to filter more reasonable text representations and built a matching matrix at different granularities. Then, the convolution neural network (CNN) was used to capture the matching information in each granularity. Finally, the captured matching features were input into the fully connected neural network to obtain the matching similarity. By making some experiments, the results indicate that the MG-CMF model not only gets multiple granularity representations of sentences but also can obtain matching information from multiple granularities of sentences better than the other text matching models.     

WDM-PON技术在5G前传承载中的应用研究

前传在5G(第五代通信技术)承载网性能与网络优势发挥上起到越来越重要的支撑作用。文章通过各类技术在5G前传承载中的应用研究,充分对比差异及优缺点,得出C-RAN(集中化无线接入网)方式下高速WDM-PON(波分复用无源光网络)作为前传技术具备建网快捷、建设成本低、运营维护简便等多种优势,为WDM-PON技术在5G前传网络中的规模部署提供有力支撑,也能更快推动WDM-PON产业链的成熟和发展。     

Ultrafine porous boron nitride nanofiber-toughened WC composites

Porous fibers are widely used in catalysis chemistry and hydrogen storage but are rarely used in structural ceramics. In this study, spark plasma sintering was used to prepare an ultrafine porous boron nitride nanofiber-toughened WC composite for the first time. The obtained WC-0.05 wt% ultrafine porous boron nitride nanofiber composites exhibited better properties (ie, a 2.3% increase in hardness and a 19.6% increase in fracture toughness) compared to those of the pure WC specimen. The fiber porosity improved the second phase-WC matrix microstructural combination. The described approach is a novel preparation method for the WC composites. Furthermore, a new toughening mechanism, which is based on “pinning and stretching”, was determined. These findings suggest that porous boron nitride fibers can be considered to be second phases for toughening the WC composites.     

某含锌锡多金属硫化矿石锌浮选试验研究

某含锌锡多金属硫化矿石Zn、Sn、Fe、S含量分别为6.04%、1.05%、29.33%、19.08%，锌主要以铁闪锌矿、闪锌矿的形式存在，锡主要以锡石的形式存在，铁主要以黄铁矿、磁黄铁矿等形式存在，其中的金属矿物共生关系密切，相互包裹现象普遍。为确保不影响后续选锡，对锌浮选流程进行了试验研究。结果表明：矿石在磨矿细度为-0.074 mm占80%的情况下，采用预先脱硫—锌硫混浮再分离流程处理，在选择硫酸铜为锌矿物活化剂、丁基黄药为捕收剂、松醇油为起泡剂、石灰为黄铁矿抑制剂的情况下，经1段脱硫、2粗1扫锌硫混浮、1粗2精锌硫分离，锌硫分离精选尾矿与锌硫混浮扫选精矿2次精选后锌硫分离，最终获得锌精矿Zn品位47.06%、回收率90.76%，试验指标良好。