GD2308高速加固缝纫机压脚有限元建模与分析

以工业缝纫机的减振降噪为应用背景,针对其结构特点及工作特性,通过CATIA对GD2308高速加固缝纫机压脚进行有限元建模,应用有限元软件ANSYS对压脚进行网格的划分,生成有限元分析模型,对压脚进行动态特性分析,计算压脚的前10阶固有频率与前6阶振型,给出振型图.分析结果表明,压脚结构振动主要表现为弯曲、扭转和弯扭组合...     

船舶辐射噪声远场空间指向性

为了获得浅海波导条件下船舶辐射噪声的远场空间指向性,提出了一种由水听器获得的船舶辐射噪声近场声压数据,推算其远场辐射声压的方法,即由近场声压数据,反演出船舶辐射噪声源模型中的多极子强度和位置,然后利用简正波理论,计算该反演参数所对应的远场指向性.仿真计算过程中,将该方法获得的远场指向性与原假设声源直接计算的远场指向性进行对比.结果表明,该方法能有效的计算出船舶辐射噪声的远场空间指向性.     

数字光接收机性能指标的优化设计

首次提出一种新型数字光接收机模型,改进了其中均衡器的设计;以S.D.Personick理论为基础,推导并计算了此光接收机基于"最坏情况"下判决时刻的噪声平均功率。研究结果表明,此模型在输入脉冲展宽较为严重的情况下表现出了良好的噪声性能;进而提高了数字光接收机的灵敏度,优化了数字光纤传输系统的误码性能;为改善数字光接收机性能指标提出了新的途径。     

红外热像视频的细微变化放大

目的 针对红外热像视频对比度低、成像模糊和难以进行细节观测的缺点,提出一种基于欧拉视角的红外热像视频细微变化放大方法。该方法可以将红外热像视频中细微的色彩变化和动作变化进行放大,将原本人眼无法察觉到的变化清晰地展示出来。方法 该方法首先采用对比度金字塔算法对红外热像视频中每一帧图像进行空域分解,其次对各个尺度的图像进行时域滤波,选择出感兴趣的变化频率并进行线性放大,然后对放大后的信号进行重构,最后对重构得到的图像进行降噪处理,从而获得细微变化放大的红外视频。结果 针对色彩放大和动作放大,实验采集了若干红外热像视频。其中,对人脸侧面的颜色进行放大时,选择像素值变化频率在0.751 Hz 范围内的信号进行滤波并放大,得到像素值变化被放大100倍的视频;对吉他弦的动作进行放大时,选择变化频率在100120 Hz范围内的信号进行滤波并放大,得到弦的动作幅度被放大的视频。结果表明该方法可以使视频中所选择的变化频段得到有效增强。结论 本文方法可以放大红外视频中原本无法观测到的细微变化,并使之清晰呈现,在军用和民用领域中有着广泛用途。     

The use of electrochemical noise methods (ENM) to study thick, high impedance coatings

Thick, high impedance organic coatings are those class of coatings used to provide corrosion protection to naval vessels, pipelines, gasoline storage tanks, and other large structures such as bridges and plant structures. These coatings, especially the newest generations now being used in practice, can provide exceptional protection and lifetime of performance such that properly and accurately assessing and differentiating among competing coatings is a very difficult task. The standard protocol of salt fog testing (ASTM B117), immersion testing, and outdoor exposure in a corrosive environment with subjective evaluation of a coating's performance durings and after testing, does not adequatcly rank and predict coating lifetimes for new coating systems, especially for the environmentally compliant coating systems such as powder coatings (especially the thick, fusion bonded epoxy (FBE) coatings used for pipelines), two component epoxy and urethane coatings and waterborne coatings. New, objective test methods are desperately needed by users and manufacturers of coatings. A relatively new electrochemical test procedure, electrochemical noise methods (ENM), as developed by Skerry and Eden, has been shown in our laboratory to be very successful in the ranking and prediction of relative coating performance. We have used the method successfully on naval ship coatings, several pipeline coatings and other related systems, and Skerry has used them successfully on industrial maintenance coatings. We have used these methods in conjunction with electrochemical impedance spectroscopy, d.c. resistance measurements and cyclic salt fog testing of the Prohesion^TM type. In our studies of pipeline coatings, we needed to investigate thermal effects because of their extended range of use temperature. In these studies, we have discovered that electrochemical methods can be used for an in situ measurement of the T_g of coatings in electrolyte immersion. Further, the ‘plasticizing’ effect of aqueous electrolyte absorption as well as its relative irreversibility has been shown. For all coatings studied, ENM provided useful, objective, numerical data which rapidly ranks coatings and provides useful information on the relative lifetime prediction of coatings which may provide up to 30 years of service.     

Robust sparse principal component analysis

The model for improving the robustness of sparse principal component analysis（PCA） is proposed in this paper. Instead of the l2-norm variance utilized in the conventional sparse PCA model,the proposed model maximizes the l1-norm variance,which is less sensitive to noise and outlier. To ensure sparsity,lp-norm（0 p 1） constraint,which is more general and effective than l1-norm,is considered. A simple yet efficient algorithm is developed against the proposed model. The complexity of the algorithm approximately linearly increases with both of the size and the dimensionality of the given data,which is comparable to or better than the current sparse PCA methods. The proposed algorithm is also proved to converge to a reasonable local optimum of the model. The efficiency and robustness of the algorithm is verified by a series of experiments on both synthetic and digit number image data.     

数学形态学在数字图像处理中的应用研究

传统边缘检测算法由于对噪声敏感,难以准确提取图像边缘,导致图像处理效果不佳.基于传统数学形态学算法中结构算子的方向性和尺寸几何的基础上进行算法改进.针对抗噪型碰撞腐蚀形态学边缘检测算子结构元素特征,采用不同大小结构元素组合来提取边缘特征,有效保证了图像细节的同时去掉较大噪声点.根据结构元素的方向性,利用同向结构元素图像的匹配来检测各边缘信息,确保不同向边缘信息的完整度.通过比较文本改进算法与传统的边缘检测算法对图像边缘检测效果表明:本文提出的改进算子在处理较大图像边缘检测时具有更快的检测速度,且图像边缘光滑,细节清晰,具备了更强的抗噪性能.     

基于稳健设计的SGNS算法的超参数调优方法

通常给定超参数的若干取值选取性能最大的为最优组合(称为直接选优法),但是此方法的稳健性差。为此,提出了一种基于稳健设计的超参数调优方法(称为稳健调优法)。具体地,以SGNS算法中的超参数调优为例,在词语推断任务上实验并得出:经方差分析得到SGNS算法中的七个超参数中的五个对算法预测性能有显著影响,确定为主控因子,其余两个确定为噪声因子,且主控因子中有三个对性能估计的方差有显著影响,因此,调优中仅从期望最大来直接选优是不合理的;稳健调优法与直接选优法两者在预测性能上没有显著差异,但稳健调优法对噪声因子具有较好的稳健性。稳健调优法对一般的深度神经网络的调参有实际的借鉴意义。     

基于局部概率抽样的标签噪声过滤方法

分类学习任务中,在获取数据的过程中会不可避免地产生噪声,特别是标签噪声的存在不仅使得学习模型更复杂,而且容易造成过拟合并导致分类器泛化能力的下降。标签噪声过滤算法虽然在一定程度上可以解决上述问题,但是仍然存在噪声识别能力较差、分类效果不够理想以及过滤效率低等问题。针对这些问题,提出一种基于标签置信度分布的局部概率抽样方法来进行标签噪声过滤。首先利用随机森林分类器对样本的标签进行投票,从而获取每个样本的标签置信度;然后根据标签置信度的大小,将样本划分为易识别样本和难识别样本;最后分别采用不同的过滤策略对样本进行过滤。实验结果表明,在标签噪声存在的情况下,所提方法在大多数案例上能够保持较高的噪声识别能力,并且在分类泛化性能上也具有明显优势。     

广义空间调制系统中的物理层抗窃听传输方案

为实现信息的保密传输,提出一种基于广义空间调制技术的物理层安全方案.发射端根据合法信道的状态信息对信号进行预处理,使合法接收者接收到来自各激活发射天线的信号相位对齐,并对不同激活天线组合的信号进行不同的相位旋转,提高合法接收者的接收性能.而窃听者接收到的信号相位仍然为随机分布,接收性能远低于合法接收者,有效地保护通过天线索引传输的信息.发送端同时发送指向窃听者的人工噪声,保护通过幅相调制携带的信息不被窃听.对保密容量、误码性能和信号与人工噪声的功率分配进行理论分析,仿真结果表明,通过合理分配功率,合法接收者的误码性能明显优于窃听者,能获得可观的系统保密容量.