Matching ratio between positive and negative electrodes for double-layer capacitors

For double-layer capacitors in alkaline electrolyte, the specific capacitance of the positive electrode is not equal to that of the negative one. Thus, capacitor performance cannot be optimal with a positive/negative electrode matching ratio of 1. In this study nanoporous glassy carbons (NPGCs) were employed as the electrodes of capacitors, and the influence of matching ratio between positive and negative electrode on capacitor performance was systematically investigated. In aqueous KOH, the specific capacitance of the positive electrode is lower than that of the negative electrode. The matching ratio at which a maximum capacitance is obtained is dependent on the values of the positive and negative electrode capacitance. At low current rate, the highest specific capacitance is achieved at a matching ratio slightly higher than 1. At high current rate, a capacitor has the highest specific capacitance with the lowest resistance at a matching ratio of 1.5. This indicates that an optimum matching relationship between positive and negative electrodes is attained.     

Preparation of phenolic resin derived 3-D ordered macroporous carbon

Three-dimensionally ordered long-range macroporous carbon structures were prepared using commercially available phenolic resin by utilizing sacrificial colloidal silica crystalline arrays as templates that were subsequently removed by HF etching after pyrolysis in an argon atmosphere. SEM, TEM, and BET were employed to characterize the morphology and the surface area of the porous carbon structures. The pore size (150–1000 nm) and BET surface area, which reflect pore volume (298.6 m²/g (1.32 cm³/g) ∼ 93.7 m²/g (0.12 cm³/g)), of the macroporous carbon structures produced were approximately proportional to the size (150–1000 nm) of the sacrificial silica sphere templates used (annealing temp. 550°C). The achieved 550 nm porous carbon structures were examined to function as potential catalyst carriers and were successfully impregnated with Ag or Pt-Ru on their inner walls after borohydride reduction at room temperature. In addition, porous carbon patterns were fabricated using the ‘micromolding in capillary’ technique, which has potential applications in the microreaction technology.     

到达数据中时空异常聚簇发现

在时空数据中有一类表示用户在某一时间到达某一地点的数据——到达数据,到达数据可以是社交网站的签到数据、轨迹数据中的停留点及公共交通中乘客抵达的位置数据,这些数据的聚簇可以反映用户的聚集行为.基于到达数据,提出一类新的时空数据查询——时空异常聚簇发现.将到达数据进行周期性划分,通过时空聚类算法对一个时间段的数据进行聚类,比较不同时间段内聚簇的差异度,发现具有最大簇异常度的前k个簇.通过该查询发现的时空异常聚簇可以应用于城市安全管理、基于位置的服务和交通调度等方面.定义了异常簇查询模型,提出了针对任意形状聚簇的簇差异度度量,将异常簇查询转化为二分图最大匹配问题,对二分图构建与匹配进行了优化并提出了高效的查询算法.利用真实数据集进行了充分实验,验证了查询结果的实际意义,评估了所提出的各查询算法在不同参数设置下的查询效率.     

一种合成聚焦的便携式B型超声成像方法

传统合成聚焦超声成像方法在图像重建过程中处理数据量较大,致使其在便携式B型超声设备中应用受限。针对该问题,提出一种基于压缩感知理论和合成聚焦波束形成算法的便携式B型超声成像方法(SFCS)。该方法通过对接收到的射频回波信号进行随机采样得到欠采样的测量信号,将测量信号经过合成聚焦波束形成处理获得欠采样的射频信号线,基于压缩感知的信号重构算法,由欠采样的射频信号线高质量地恢复信号后供后续成像处理使用。仿真实验结果表明,SFCS可有效地解决合成聚焦超声成像过程中数据量大的问题,从而在保证较高成像质量的前提下满足便携式B型超声设备小型化、低成本的技术要求,具有工程应用价值。     

一种高效的多模式字符串匹配算法

在Fan-Su(FS)多模式字符串匹配算法基础上,结合BM-Horspool(BMH)算法和Quick Search(QS)算法的优点,提出一种高效的多模式字符串匹配算法。该算法能够充分利用本次匹配失败和部分匹配成功的信息,一方面增加模式树根节点失配的概率,提高匹配过程中失配时的跳跃距离。另一方面避免不必要的状态转移,实现不匹配时的连续跳转。分析指出,在最好情况和平均情况下,时间复杂度均优于ACBM算法和FS算法。实验结果表明,一般情况下该算法的查找时间仅为AC算法的10%~35%,ACBM算法的50%~60%,FS算法的70%左右,FSQB算法的65%左右。     

Preparation of three-dimensional ordered macroporous SiCN ceramic using sacrificing template method

Three-dimensional (3D) long range well ordered macroporous SiCN ceramics were prepared by infiltrating sacrificial colloidal silica templates with the low molecular weight preceramic polymer, polysilazane. This was followed by a thermal curing step, pyrolysis at 1250 °C in a N₂ atmosphere, and finally the removal of the templates by etching with dilute HF. The produced macroporous SiCN ceramics showed high BET surface areas (pore volume) in the range 455 m²/g (0.31 cm³/g)–250 m²/g (0.16 cm³/g) with the pore sizes of 98–578 nm, which could be tailored by controlling the sizes of the sacrificial silica spheres in the range 112–650 nm. The sphere-inversed macropores were interconnected by 50 ± 30 nm windows and 3–5 nm mesopores embedded in the porous SiCN ceramic frameworks, which resulted in a trimodal pore size distribution. The surface of the achieved porous SiCN ceramic was then modified by Pt–Ru nanoparticle depositing under mild chemical conditions.     

一种并行中英文混合多模式匹配算法

针对中英文混合文本的匹配准确性及大规模数据文本的匹配效率等问题,基于经典的线索化完全哈希特里树算法,提出一种并行化的中英文混合多模式文本匹配算法。采用拆分文本降低多模式匹配算法的串行度,进而在拆分出的小文本上并行地执行文本匹配。通过并行化预处理过程,设计新的存储结构。实验结果表明,该算法在保证结果正确的前提下,执行效率高于经典的串行匹配算法,当数据规模达到226个字符时,可以获得8倍以上的加速比。     

一种改进的视网膜图像中微小动脉瘤的检测算法

基于糖尿病性视网膜病变中最早出现的微小动脉瘤病症进行了研究,提出一种有效的微小动脉瘤检测算法。首先在传统模板匹配算法的基础上提出了一种动态多参数模板匹配算法,并且使用相对误差和与相关系数来共同制约匹配度,从而实现了更为精确的匹配提取;其次提出了基于分布特性的计分策略和自适应加权的汇总策略,避免了单纯采用各个特征量作为独立约束指标进行筛选时忽视各个特征量的约束力大小的弊端。实验结果表明,该检测算法能够有效地提高微小动脉瘤的检测真阳性率。     

快速鲁棒特征算法的CUDA加速优化

提出一种基于统一计算设备架构(Compute Unified Device Architecture,CUDA)的快速鲁棒特征(Speed-up Robust Feature,SURF)图像匹配算法。分析了SURF算法的并行性,在图像处理单元(Graphics Processing Unit,GPU)的线程映射和内存模型方面对算法的构建尺度空间、特征点提取、特征点主方向的确定、特征描述子的生成及特征匹配5个步骤进行CUDA加速优化。实验表明,相比适用于CPU的SURF算法,文中提出的适用于GPU的SURF算法在处理30MB的图片时性能提高了33倍。适用于GPU的SURF算法拓展了SURF算法在遥感等领域的快速应用,尤其是大影像的快速配准。