Cyclodextrin-induced auto-healing of hybrid polyoxometalates

You'll never walk alone: An auto-healing process of the first host-guest complex involving a polyoxometalate hybrid and a cyclodextrin is reported. Indeed, this inclusion complex allows the complete restoration of the anchored organic moiety released by a basic stress, a process otherwise not fully reversible.     

Orthogonal addressable monolayers for integrating molecular logic

Plug and play: The mimicking of integrated circuits by using two individual monolayers (molecular chips) is shown. These monolayers can be individually addressed using identical inputs. Upon combination of their optical outputs, the input/output characteristics of a molecular encoder is obtained. Since the encoder functionality is only displayed when both chips are active, the device behaves according to a plug-and-play principle (In=input; see picture).     

高加速应力筛选试验中惯性器件性能建模与一致性测试

针对高加速应力筛选(High Accelerated Stress Screening)试验中对受试产品剩余寿命的要求,在惯性器件剩余寿命降低到一定程度后其性能会发生变化的前提下,提出了一种考核惯性产品寿命受高加速应力影响的方法。首先在高加速应力筛选试验前测试惯性器件的关键性能,以信噪比(SNR)和相关度(ρ)为评价原则利用自回归(Auto Regressive)方法对性能指标测试参数进行建模,确定最优模型阶数;使用Burg方法计算AR模型系数。在惯性器件经受了高加速应力筛选试验后测试惯性器件的性能参数,利用Chow检验方法对加速应力筛选试验前后的性能测试参数进行检验,在一定置信度条件下考察性能参数是否发生了显著性变化,进而对高加速应力筛选试验是否影响惯性器件的剩余寿命进行判定。最后通过某型加速度计高加速应力试验进行了验证。     

Study on the Preparation of Cr‐based Catalysts Doped by Zn using Sol–Gel Auto‐Combustion Method and Its Application for Synthesis of 1‐Chloro‐2,3,3,4,4,5,5‐heptafluorocyclopentene

High‐surface‐area chromium‐based catalysts in the presence of a small amount of zinc were prepared via a sol–gel auto‐combustion method using chromic nitrate, zinc nitrate, and citric acid. First, the auto‐combustion behavior of the dried gel was investigated by derivative thermogravimetry and (DTG)‐TG and infrared (IR) techniques. The results revealed that the dried gel exhibited self‐propagating combustion properties. Second, the as‐burnt powders were characterized by IR , X‐ray diffraction (XRD) , Brunauer–Emmett–Teller analysis (BET) , and scanning electron microscopy (SEM) . The findings showed that the gels were directly converted into CrZn ‐O nanoparticles with high surface area during the auto‐combustion process. Third, the pre‐fluorination Cr‐Zn catalysts were characterized by XRD , BET , SEM , X‐ray photoelectron spectroscopy (XPS) , and Fourier transform (FT)‐IR spectroscopy of pyridine adsorption techniques. It was found that the presence of zinc led to significant structural changes in the catalyst, the particle size was smaller, the surface area became larger, and more active sites appeared. Finally, the catalytic activities of the samples were tested for the fluorination of 1,2‐dichlorohexafluorocyclopentene (1,2‐F6 ) with anhydrous hydrogen fluoride. The obtained results indicated that the pre‐fluorination activated Cr‐Zn catalysts prepared by this sol–gel auto‐combustion method exhibited high efficiency in the synthesis of cyclic hydrofluorocarbons.     

基于坐标旋转数字计算算法的小型光电编码器细分

为在不增加体积的前提下提高小型光电编码器精度,分析了计算法细分误差产生的原因,提出了基于坐标旋转数字计算(CORDIC)算法的光电编码器精码信号新细分法,利用简单的移位和加法操作可实现对采集到的正交码盘精码信号直接细分求相位,避免了查"细分表"引入的细分误差。对细分算法进行了分析与优化,使算法在取得合适精度的同时提高了运算速度。运用研究的细分法对某16位小型光电编码器精码信号进行256份细分时,比利用计算法细分时编码器的均方根误差减小了一半。实验结果表明,研究的新细分法可直接对光电编码器精码信号进行高精度细分,对于研制小型化、高精度光电编码器具有重要意义。     

准绝对式光学编码器

准绝对式编码器继承了绝对式和增量式光学编码器的优点,根据串行位置编码的基本原理,借助计算机辅助设计,获得了全周100个位置的编码方案,设计了相应的准绝对式编码器;通过与两种传统编码器特点的比较,突出了准绝对式光学编码器的显著特点,指明了其未来良好的应用和发展前景.     

汽车消费、住房抵押贷款违约特性及对策

本通过对比住房抵押贷款和汽车消费贷款的违约特性，得出在利率由央行统一规定的条件下，对住房抵押贷款和汽车消费贷款，实行同样的首付款政策是不合理的，这导致目前汽车消费贷款的违约率居高不下。对于住房抵押贷款，银行可以适当降低首付款，来提高本银行住房抵押贷款在市场的竞争力，对于汽车消费贷款，可以通过采取提高汽车贷款首付款的措施，来降低违约率，控制抵押贷款风险。     

基于支持向量机的核磁共振左心室图像自动检测与分割

提出了用基于支持向量机的方法来实现核磁共振左心室图像的自动分割方法。首先用经过训练的支持向量机(SVM)在二维图像中进行识别和定位左心室目标区域并进一步找出边缘区域，采用一种改进的训练方法来提高SVM识别率，然后在足够准确的区域中利用梯度方法找出边缘点，并把他们连接起来，找出目标的边缘，达到分割的目的。实验表明，这种分割方法降低了SVM对背景图像的敏感度，提高了SVM识别率。     

基于深度自动编码器的机场安检人脸识别系统设计

由于现有的机场安检系统通常只针对行李和旅客携带的违禁物品进行检测,没有考虑到对于正在通缉的罪犯进行监控和核查,设计了一种基于Gabor小波滤波和深度自动编码器的机场安检人脸识别系统。首先,以ATmega128L为处理器核心,SA7111作为模数转换器,MAX7000作为全局逻辑控制单元设计了机场人脸识别系统硬件,然后采用Gabor小波函数作为卷积核函数,在对原始图像分块的基础上进行卷积,采用多个RBM堆叠组成的自动编码器,通过比较差异算法训练RBM从而自动提取人脸特征,最后构建一个三层的BP神经网络,将自动提取的人脸特征作为输入,将图像标签作为输出层,并通过反向传播算法训练网络进行人脸识别。通过部署仿真实验环境对文中方法进行验证,仿真结果表明：文中系统能较为精确地实现人脸识别,与其它方法相比,具有识别率高和收敛速度快的优点。