Using attention LSGB network for facial expression recognition

Pattern Analysis and Applications - Both the multiple sources of the available in-the-wild datasets and noisy information of images lead to huge challenges for discriminating subtle distinctions...     

海洋钢结构尺寸误差分析新方法

海洋钢结构精度是反映钢结构建造的一项重要指标,传统海洋钢结构精度分析方法是通过对比实际测量尺寸与理论尺寸的差值反映钢结构的误差,这种方式简单却不够直观,不利于施工人员进行尺寸的调整。运用最小二乘坐标拟合技术,使实测数据最优匹配设计数据,并使钢结构的尺寸误差以三维偏差的大小和方向表示,直观合理,并利用AutoCAD VBA编写计算程序,快速计算出偏差值,对现场施工调整及误差评估具有很好的作用。     

时空域波动方程混合网格有限差分数值模拟

传统2 M阶有限差分格式（Tradtional 2 Mth-order Finite-Difference Schemes,T2 M-FD）和时空域2 M阶有限差分格式（Time-Space Domain 2 Mth-order Finite-difference,TS2 M-FD）均是目前应用较普遍且具代表性的高精度有限差分方法。T2 M-FD仅基于空间域频散关系求解差分系数,模拟精度较低。TS2 M-FD基于时空域频散关系和平面波理论求解差分系数,模拟精度较高。T2 M-FD和TS2 MFD的差分格式相同,都是只利用常规直角坐标系中坐标轴上的网格点差分近似波动方程中的Laplace算子,而没能充分利用旋转直角坐标系中距离差分中心点更近的网格点来进一步提高模拟精度。本次研究提出利用常规直角坐标系和旋转直角坐标系中的网格点一起差分近似波动方程中的Laplace算子,并将Laplace算子表示为常规直角坐标系中M个Laplace算子和旋转直角坐标系中N个Laplace算子的加权平均,构建出一种新的混合2 M+N型有限差分格式（M2 M+N-FD）。推导出M2 M+N-FD基于时空域频散关系和平面波理论的差分系数计算方法,进行频散及稳定性分析。频散分析表明：与T2 M-FD和TS2 M-FD相比,M2 M+N-FD能更有效地压制数值频散,模拟精度更高。稳定性分析表明：M2 M+N-FD和TS2 M-FD的稳定性基本相当,比T2 M-FD的稳定性强。最后,利用M2 M+N-FD进行均匀介质和层状介质模型的数值模拟试验,并将其推广应用于Marmousi模型的逆时偏移,高精度的数值模拟结果和偏移成像质量证明了M2 M+N-FD的优越性和普遍适用性。     

中美导淮借款的评析

张謇在民初曾力图借用外资治理淮河.美国红十字会在美国政府鼓励和支持下,在考察后与中方商谈并签署了<导淮借款草约>.但由于中国国内局势动荡和第一次世界大战爆发,中美一直未能签订导淮借款正式合同,美国参与、控制导淮工程的计划遂告失败.对中关导淮借款进行分析,有助于认识张謇水利经济思想的历史价值,其利弊得失可以为现今水利工程筹集资金模式的完善提供一些借鉴.     

Morphological structure and mechanical properties of In situ microfibrillar composites of modified PA66 with PP

In situ microfibrillar composites (PP/mPA66) of modified polyamide66 (mPA66) with polypropylene (PP) were prepared by using a “post‐compatibilization” technique. The mPA66 was firstly obtained by reactive extrusion of PA66 resin with a specially designed compatibilizer, which was then blended with PP through extrusion combined with a hot stretching and subsequently quenching process. The PP/mPA66 in situ microfibrillar composites were comparatively studied with simply blended samples of PP/PA66 that were prepared by blending PA66 and PP together with (or without) the same compatibilizer through extrusion. PA66‐g‐PP (and/or elastomers) graft copolymer formation in mPA66 was identified by dissolution test and infrared spectroscopy measurement, the compatibilizer is unevenly dispersed with large domains in PA66 as observed by scanning electron microscope (SEM). In PP/mPA66 composites, the in situ generated PA66 microfibrils have a rather nonuniform diameter distribution and a very rough surface. SEM observations for the fractured surface illustrated that PP/mPA66 composites have structural characteristics of stronger adhesion and moderate flexibility of the interface. Enhanced compatibilization between the PA66 microfibrils with the PP matrix resulted in improved mechanical properties of the PP/mPA66 composites. With optimized composition, the PP/mPA66 composite has notched Izod impact strength, flexural modulus, and tensile yield stress of 1.49, 1.16, and 0.99 times as those of the neat PP, respectively. Such enhanced mechanical properties balance and improved interface adhesion were not found in the simply blended samples of PP/PA66 with or without the specially designed compatibilizer. © 2012 Wiley Periodicals, Inc. J. Appl. Polym. Sci., 2013     

MEMS镍平面弹簧非线性力学性能仿真及研究

工程应用中,平面微弹簧存在弹性系数的非线性现象且难以被忽略,弹簧结构的几何非线性是引起这种现象的主要原因。针对特定形状的弹簧运用Ansys软件进行了非线性有限元静力学仿真,计算出弹簧弹性力F与弹性形变x之间的对应关系,并与采用MEMS微细加工技术制得弹簧样品实测实验结果比较,相对误差低于3.2%,验证了分析模型的可靠性。而后设计了5种不同平面形状的微弹簧,仿真计算它们的力学性能曲线,以及改变弹簧臂宽度厚度后的力学性能曲线,分别进行比较以讨论影响非线性的主要因素,为弹簧的设计工作提供参考依据。     

雾环境下非视距日盲紫外光通信链路路径损耗研究

利用Luettgen单次散射信道模型对雾环境下非视距日盲紫外光通信信道传输特性进行了模拟分析.选取波长为260 nm的紫外光,分析了雾环境下对于不同的通信距离,该紫外光的路径损耗与能见度的关系.结果表明:在通信距离已知以及系统的发射仰角、接收仰角、发射视场半角和接收视场半角等几何参数已知的情况下,存在一个使得紫外光传输的路径损耗达到最小的能见度;同时,分析并比较了0～2 km辐射雾和平流雾传输路径损耗与能见度、通信距离的关系,并得到辐射雾环境更利于紫外光传输.     

基于DRAM牺牲Cache的异构内存页迁移机制

当海量数据请求访问异构内存系统时,异构内存页在动态随机存储器(dynamic random access memory,DRAM)和非易失性存储器(non-volatile memory,NVM)之间进行频繁的往返迁移.然而,应用于传统内存页的迁移策略难以适应内存页"冷""热"度的快速动态变化,这使得从DRAM迁移至N...     

Content-based deep communication control for networked control system

In smart cities, the networked control system plays a significant role in transportation systems, power stations or other critical infrastructures, and it is facing many security issues. From this point, this paper proposes a content-based deep communication control approach to guarantee its security. Based on the layer architecture, this approach analyzes the interactive content in depth according to different industrial communication protocols, and implements the access control between two distinct enclaves. For OPC Classic, we acquire the dynamic port provided by OPC server, and open a new connection belonging to this port; for Modbus/TCP, we not only analyze the ordinary function codes and addresses, but also check the register or coil values by using the multi-bit Trie-tree matching algorithm. Besides, the white-listing strategy is introduced to satisfy the special requirements of industrial communication. Our experiment results show that, on the one hand the proposed approach provides OPC and Modbus/TCP defenses in depth; on the other hand it has less than 1 ms forwarding latency and 0 packet loss rate when the rule number reaches 200, and all these meet the availability requirements in the networked control system. In particular, this approach has been successfully applied in several real-world petrochemical control systems.