Ensemble relation network with multi-level measure

Fine-grained few-shot learning is a difficult task in image classification. The reason is that the discriminative features of fine-grained images are often located in local areas of the image, while most of the existing few-shotlearning image classification methods only use top-level features and adopt a single measure. In that way, the localfeatures of the sample cannot be learned well. In response to this problem, ensemble relation network with multi-level measure (ERN-MM) is proposed in this paper. It adds the relation modules in the shallow feature space tocompare the similarity between the samples in the local features, and finally integrates the similarity scores from thefeature spaces to assign the label of the query samples. So the proposed method ERN-MM can use local details andglobal information of different grains. Experimental results on different fine-grained datasets show that the proposedmethod achieves good classification performance and also proves its rationality.     

基于LabVIEW的惯性测量单元信号采集及处理系统设计

利用LabVIEW设计了惯性测量单元信号采集及处理软件,主要包括串口通信、数据分离及信号处理等模块;串口通信模块采用VISA方式编程,建立了惯性测量单元与上位机之间的通信;根据惯性测量单元输出数据的格式,利用LabVIEW中的匹配模式函数分离出沿三轴向的加速度信号和角速率信号;设计了一种用于去除加速度信号中低频成份的数字滤波器;在建立的陀螺漂移模型基础上,分别进行小波基、小波分解层数、阈值函数及阈值估计方法等小波参数的选取.     

轨道空间线形的运动检测算法

基于惯性测量平台,对惯性测量单元(IMU)信息进行等空间采样.采用经验模态分解算法(EMD)消除信号中的干扰成分,设计数字积分器对惯性信息进行积分计算,获取相对空间运动轨迹.实验结果表明:上述方法所测量的轨道三轴绝对位移精确度大于97.17%.     

仿真软件在计算机网络通信教学中的应用

基于PacketTracer仿真软件组建了跨路由器的虚拟VLAN网络实验,分析了路由器的端口设置、IP地址分配以及交换机的详细配置方法,丰富了实验教学手段。通过具体的实验对虚拟网络进行了仿真测试分析,实现了跨路由器间的VLAN通信以及跨网络的DHCP中继服务,并在实际的网络设备上进行了试验验证。     

基于刚柔耦合的转向架构架关键部位疲劳分析

以某型地铁车辆转向架构架为研究对象,根据结构模态综合法基本原理,通过有限元模态分析对构架进行主自由度缩减,得到构架模态中性文件。建立地铁车辆刚柔耦合多体动力学模型,利用多体动力学仿真,得到不同工况下构架关键部位的应力时间历程变化趋势。最后采用惯性释放法对构架关键部位进行准静态应力分析,依据车辆刚柔耦合多体动力学模型得到转向架构架关键部位的边界载荷,结合材料的S-N曲线,最后通过虚拟疲劳仿真的方法预测构架关键部位的疲劳寿命。     

软土地层并行曲线隧道施工顺序对既有隧道的影响

为了研究并行曲线隧道施工顺序对既有隧道的影响,以横琴杧洲隧道工程为依托,建立了并行曲线隧道三维有限元数值模型。在此基础上,研究了不同施工顺序和曲率半径r(r=500,800 m)下新建曲线盾构隧道开挖对既有隧道变形的影响。结果表明:施工顺序对既有隧道位移影响较小,曲率半径对既有隧道位移影响相对较大; 随着曲率半径(r=500～800 m)的增加,既有隧道位移增加约15%; 既有隧道的位移主要在盾构开挖面前方2D(D为隧道外径)、后方1D范围内产生; 施工顺序对既有隧道内力的影响与曲率半径有关; 隧道曲率半径为500 m时,施工顺序对既有隧道内力变化影响规律相似,即离盾构开挖面最近的既有隧道剖面产生的弯矩最大,且最大弯矩和最小弯矩均出现在靠近新隧道一侧; 内侧隧道先开挖时,既有隧道的弯矩(绝对值)更小,此时对于并行曲线隧道施工,内侧隧道先开挖更安全; 在盾构开挖面前方一定距离内既有隧道产生的轴力最大; 隧道曲率半径为800 m时,双线隧道近似于平行隧道,施工顺序对既有隧道内力变化和大小影响较小。     

基于SCWS的SIM卡门户网站的实现

根据国际电信市场的经验,要更有效地推广3G业务,移动运营商应采取与2G时代不同的全新的促销方法。SIM卡门户网站采用SCWS技术,将Internet、移动网络、移动服务器平台结合起来,使用手机上的浏览器,帮助用户发现并使用运营商的数据业务,引导最终用户购买该项数据业务。文中从业务分析入手,介绍了SCWS技术的SIM卡门户网站应用环境、工作流程、通信协议,为电信运营商、手机厂商和SIM卡供应商提供参考。     

Influence of FRP wrapping techniques on the compressive behavior of concrete prisms

Results of an experimental investigation on the compressive behavior of concrete prisms with square cross-section, externally wrapped with carbon fiber reinforced plastic (CFRP) sheets are presented. The effects of the following parameters were analyzed: local reinforcement at the corners and continuous layers; horizontal and vertical discontinuous strips; number of continuous layers—length of specimens.An analytical model is proposed to determine the maximum bearing capacity of compressed concrete members with square cross-section and externally wrapped with FRP for the different configurations examined, and also able to consider the strength reduction with the length increase of concrete members. Analytical results are then compared with the experimental data available in the literature, showing good agreement.     

桥面横向连接对双幅桥梁抗风性能的影响

以泉州海湾桥两种结构形式双幅桥面主梁方案为背景,通过数值计算研究了两种双幅桥面结构的结构动力特性;通过风洞试验研究了桥梁颤振稳定性和涡激共振性能。研究结果表明,是否存在桥面横向连接对双幅桥的结构固有动力特性影响较大,无横向连接双幅桥面较有横向连接双幅桥面的模态特征更复杂,其竖弯和扭转基频均会出现双幅桥面同向运动和反向运动两种模态;在气流作用下,无横向连接双幅桥面和有横向连接双幅桥面的主梁运动形式不同,两种结构形式双幅桥面的颤振稳定性能和涡激共振性能存在显著差异,双幅有横向连接主梁的气动稳定性优于双幅无横向连接主梁,但前者的涡激共振性能要差于后者。     

NOISE IDENTIFICATION FOR HYDRAULIC AXIAL PISTON PUMP BASED ON ARTIFICIAL NEURAL NETWORKS

The noise identification model of the neural networks is established for the 63SCY14-1B hydraulic axial piston pump. Taking four kinds of different port plates as instances, the noise identification is successfully carried out for hydraulic axial piston pump based on experiments with the MATLAB and the toolbox of neural networks. The operating pressure, the flow rate of hydraulic axial piston pump, the temperature of hydraulic oil, and bulk modulus of hydraulic oil are the main parameters having influences on the noise of hydraulic axial piston pump. These four parameters are used as inputs of neural networks, and experimental data of the noise are used as outputs of neural networks. Error of noise identification is less than 1% after the neural networks have been trained. The results show that the noise identification of hydraulic axial piston pump is feasible and reliable by using artificial neural networks. The method of noise identification with neural networks is also creative one of noise theoretical research for hydraulic axial piston pump.