Simultaneously achieved high-energy storage density and efficiency in (K,Na)NbO3-based lead-free ferroelectric films

With the development of advanced electrical and electronic devices and the requirement of environmental protection, lead-free dielectric capacitors with excellent energy storage performance have aroused great attention. However, it is a great challenge to achieve both large energy storage density and high efficiency simultaneously in dielectric capacitors. This work investigates the energy storage performance of sol-gel-processed (K,Na)NbO₃-based lead-free ferroelectric films on silicon substrates with compositions of 0.95(K_0.49Na_0.49Li_0.02)(Nb_0.8Ta_0.2)O₃-0.05CaZrO₃-x mol% Mn (KNN-LT-CZ5-x mol% Mn). The appropriate amount of Mn-doping facilitates the coexistence of orthorhombic and tetragonal phases, suppresses the leakage current, and considerably enhances the breakdown strengths of KNN-LT-CZ5 films. Consequently, large recoverable energy storage density up to 64.6 J cm⁻³ with a high efficiency of 84.6% under an electric field of 3080 kV cm⁻¹ are achieved in KNN-LT-CZ5-5 mol% Mn film. This, to the best of our knowledge, is superior to the majority of both the lead-based and lead-free films on silicon substrates and thus demonstrates great potentials of (K,Na)NbO₃-based lead-free films as dielectric energy storage materials.     

基于信息熵与SIFT算法的天文图像配准

天文图像配准是研究天体运动的一项关键技术,图像内部结构往往存在轻微的不规则运动.但是图像配准涉及到计算整个图像的变换关系,在此情况下,无论是采用基于统计特征还是基于局部特征的配准方法,都难以取得理想的效果.为此,提出基于信息熵与SIFT算法的天文图像配准方法.该方法首先需对图像进行均匀分块并计算每块熵值,以熵值最大者作为配准的局部子图,然后通过尺度不变特征变换(Scale Invariant Feature Transform,SIFT)及仿射变换建立变换关系,继而利用局部子图变换关系完成图像的配准.该方法一方面能缩短变换关系的建立时间,另一方面能保证图像中信息熵最大区域配准,有效提高天文图像配准质量.     

一种Android终端控制物联网结点的技术应用

提出了一种对楼宇分体空调进行集中管理的方法,结合物联网的构成及实际应用模式,探索并将其理论方法引入到系统中来,构建了一个基于物联网的分体空调远程集中控制系统。通过Android终端设备,可对系统运行情况进行实时监控和管理。经实验测试验证,该集中控制节点及感知节点的设计满足系统的要求,可靠性高,成本低,功能强。     

Multiphoton fluorescent images with a spatially varying background signal: a ML deconvolution method

By means of multiphoton laser scanning microscopy, neuroscientists can look inside the brain deeper than has ever been possible before. Multiphoton fluorescent images, as all optical images, suffer from degradation caused by a variety of sources (e.g. light dispersion and absorption in the tissue, laser fluctuations, spurious photodetection and staining deficiency). From a modelling perspective, such degradations can be considered the sum of stochastic noise and a background signal. Among the methods proposed in the literature to perform image deconvolution in either confocal or multiphoton fluorescent microscopy, Vicidomini et al. (2009) were the first to incorporate models for noise (a Poisson process) and background signal (spatially constant) in the context of regularized inverse problems. Unfortunately, the so-called split-gradient deconvolution method (SGM) they used did not consider possible spatial variations in the background signal. In this paper, we extend the SGM by adding a maximum-likelihood estimation step for the determination of a spatially varying background signal. We demonstrate that the assumption of a constant background is not always valid in multiphoton laser microscopy and by using synthetic and actual multiphoton fluorescent images, we evaluate the face of validity of the proposed method, and compare its accuracy with the previously introduced SGM algorithm.     

A Numerical Study of Diagonally Split Runge–Kutta Methods for PDEs with Discontinuities

Diagonally split Runge–Kutta (DSRK) time discretization methods are a class of implicit time-stepping schemes which offer both high-order convergence and a form of nonlinear stability known as unconditional contractivity. This combination is not possible within the classes of Runge–Kutta or linear multistep methods and therefore appears promising for the strong stability preserving (SSP) time-stepping community which is generally concerned with computing oscillation-free numerical solutions of PDEs. Using a variety of numerical test problems, we show that although second- and third-order unconditionally contractive DSRK methods do preserve the strong stability property for all time step-sizes, they suffer from order reduction at large step-sizes. Indeed, for time-steps larger than those typically chosen for explicit methods, these DSRK methods behave like first-order implicit methods. This is unfortunate, because it is precisely to allow a large time-step that we choose to use implicit methods. These results suggest that unconditionally contractive DSRK methods are limited in usefulness as they are unable to compete with either the first-order backward Euler method for large step-sizes or with Crank–Nicolson or high-order explicit SSP Runge–Kutta methods for smaller step-sizes. We also present stage order conditions for DSRK methods and show that the observed order reduction is associated with the necessarily low stage order of the unconditionally contractive DSRK methods. The work of C.B. Macdonald was partially supported by an NSERC Canada PGS-D scholarship, a grant from NSERC Canada, and a scholarship from the Pacific Institute for the Mathematical Sciences (PIMS). The work of S. Gottlieb was supported by AFOSR grant number FA9550-06-1-0255. The work of S.J. Ruuth was partially supported by a grant from NSERC Canada.     

基于全文检索的Segmenter分词算法改进

流行的Segmenter分词算法是开放的java源代码,它是一个很好的分词工具。但是,在全文检索系统中对分词有特殊的要求,例如歧义包客、专业词典等。所以在这里将Segmenter词典和算法都进行了改选,建立了树状词典,在匹配算法中使用了歧义包容法则、分支处理、动态规划和词库预载入。以适应电子商务案例全文检索系统。     

用改进的竞争Hopfield神经网络求解多边形近似问题

多边形近似是提取曲线特征点和简化曲线描述的一种重要方法.提出一种改进的Hopfield神经网络多边形近似算法,该算法利用选择拐点策略减少了搜索空间,重新定义了神经网络的能量函数,使其更能反映优化目标;引?入合并拆分搜索策略,有效帮助神经网络脱离局部最小值.实验结果表明,提出的改进算法是有效的,比其它算法如关键点检测法、竞争Hopfield神经网络、混沌Hopfield神经网络、遗传算法等具有更优的性能.     

Ultrahigh Energy Storage Capacitors Based on Freestanding Single-Crystalline Antiferroelectric Membrane/PVDF Composites

Inorganic/organic dielectric composites are very attractive for high energy density electrostatic capacitors. Usually, linear dielectric and ferroelectric materials are chosen as inorganic fillers to improve energy storage performance. Antiferroelectric (AFE) materials, especially single-crystalline AFE oxides, have relatively high efficiency and higher density than linear dielectrics or ferroelectrics. However, adding single-crystalline AFE oxides into polymers to construct composite with improved energy storage performance remains elusive. In this study, high-quality freestanding single-crystalline PbZrO₃ membranes are obtained by a water-soluble sacrificial layer method. They exhibit classic AFE behavior and then 2D–2D type PbZrO₃/PVDF composites with the different film thicknesses of PbZrO₃ (0.1-0.4 µm) is constructed. Their dielectric properties and polarization response improve significantly as compared to pure PVDF and are optimized in the PbZrO₃(0.3 µm)/PVDF composite. Consequently, a record-high energy density of 43.3 J cm⁻³ is achieved at a large breakdown strength of 750 MV m⁻¹. Phase-field simulation indicates that inserting PbZrO₃ membranes effectively reduces the breakdown path. Single-crystalline AFE oxide membranes will be useful fillers for composite-based high-power capacitors.     

基于多尺度分割的图像识别残差网络研究

深度卷积神经网络能够解决复杂的计算机视觉问题,被广泛应用于图像识别任务中。在基于深度卷积神经网络的图像识别过程中,增加网络的深度和宽度能够产生丰富的特征信息,使用多尺度分割方法能够有效减少冗余的特征信息。然而,增加网络的深度和进行多尺度分割都会影响识别速度。如何在保证精度的同时提高识别速度,成为设计高效网络的关键问题。通过增加网络宽度的方法对ResNet残差网络进行改进,在保证精度的基础上提升识别速度。使用ResNet-D中的残差结构并减少网络长度,得到长度只有7层的残差网络,同时对HS-ResNet中的多尺度分割方法进行优化,只保留最后一次连接合并操作,得到图像识别残差网络SSRNet。在CIFAR 10和CIFAR 100数据集上的实验结果显示,SSRNet速度最高较ResNet网络提升7倍多,同时错误率最高下降8.81%,表明缩短网络长度可大幅加快图像识别速度,同时结合多尺度分割方法能够有效提升识别精度。     

短切碳纤维增强碳化硅陶瓷基复合材料的高温冲击压缩力学性能

采用先驱体浸渍裂解法(即PIP法)制备出3种不同短切碳纤维(C_(sf))体积分数的圆柱形短切碳纤维增强陶瓷基复合材料(C_(sf)/SiC复合材料)试件,通过高温加热装置和自组装功能的霍普金森压杆装置对试件进行高温和动态荷载耦合作用下的冲击压缩试验,并通过扫描电子显微镜(SEM)观察C_(sf)/SiC复合材料断口形貌,对试件的破坏形态进行分析。试验结果表明,采用PIP法制备的C_(sf)/SiC复合材料试件中C_(sf)分布均匀,在外应力作用下C_(sf)/SiC复合材料试件发生破坏,碳纤维和碳纤维束与SiC基体脱粘被不断拔出。试件的抗压强度随C_(sf)体积含量的增加而呈现先增加后减小的变化趋势,C_(sf)体积含量为21%的试件抗压强度最高,为96.55 MPa。与常温相比,在高温压缩试验中随着复合材料试件平均温度的升高,C_(sf)/SiC复合材料试件破碎后的块度越来越大,整体性越来越好,当温度达到300℃时,C_(sf)体积含量对C_(sf)/SiC复合材料试件抗压强度的影响较小。