加权条件熵在异常检测中的应用

通过分析信息熵在异常检测中的应用和条件熵的固有特性, 提出了一种基于条件熵和加权熵理论的异常检测方法。分析网络流量特征, 选取具有内在相关性的网络特征属性, 给出了加权系数矩阵的公式和异常检测的具体实现方法。通过实验验证了该方法的有效性和实用性, 根据不同属性特征在加权条件熵理论中的应用效果, 分析了不同特征属性选择造成检测效果差异的原因。     

高速以太网中实时相似性匹配算法的研究

在高速以太网中, 针对用户行为的特殊性及数据流五元组的特点, 提出了一种自适应相似性匹配算法(ASMA-HN)。该算法充分考虑了数据流五元组的随机性分布特点, 通过一种改进的XOR-SHIFT算法将数据流分类索引, 并提出了同类数据库(SD)及异类数据库(HD)的概念, 基于距离测度分别设计了加权相似性比较函数, 使得特征值能更精准地匹配到符合要求的数据文件, 提高了整体的匹配效率。通过对校园网实际流量的现场测试, 证明该算法具有较高的精准性及可适用性, 且匹配时间较低, 能很好地满足高速网络中相似性匹配业务的需求。     

Membrane degumming of crude rice bran oil: Pilot plant study

The procedure for the classical chemical refining of vegetable oils consists of degumming, alkali neutralization, bleaching, and deodorization. Conventional refining of rice bran oil using alkali gives oil of acceptable quality, but the refining losses are very high. A critical work has been carried out to study the application of membrane technology in the pretreatment of crude rice bran oil. Oils intended for physical refining should have a low phosphorus content, and this is not readily achievable by the conventional acid/water degumming process. The application of membrane technology for the pretreatment of rice bran oil has been investigated. The process has already been successfully applied to other vegetable oils. Ceramic membranes, which are important from the commercial point of view, were examined for this purpose. The results showed that the membrane‐filtered oils met the requirements of physical refining, with a substantial reduction in color. It was observed that most of the waxy material was also rejected. Experiments were carried out to establish the relationship between permeate flux and rejection with membrane pore size, trans‐membrane pressure and micellar solute concentration.     

半参数广义线性模型惩罚似然估计的几个性质

本文在给出了半参数广义线性模型惩罚似然估计及惩罚准似然估计的基础上,得到了下面几个性质:(Ⅰ)极小惩罚似然估计及惩罚准似然估计等价于惩罚加权最小二乘估计;(Ⅱ)估计的收敛性及加速收敛法;(Ⅲ)估计的一种 Bayes 解释。     

Zernike矩结合Sigmoid曲线拟合边缘检测方法

为了改进Zernike矩边缘检测方法对非直线边缘检测精度低的弱点,提出了一种改进的基于Zernike矩方法和Sigmoid拟合法的边缘检测方法。用Zernike矩边缘检测方法检测边缘,然后用Sigmoid拟合方法检测边缘,根据边缘类型对两种方法检测的边缘结果进行加权求和,所得结果为边缘位置。分别利用理想图像和实际工业CT图像进行了实验验证,实验结果表明对非直线边缘的检测,该方法检测精度相对于Zernike矩方法提高了一个数量级。     

一种在扩散加权图像降噪中的算法

扩散加权图像中的噪声为莱斯噪声并且图像本身含有丰富的边界信息,因而要求对DWI图像有效降噪的同时,能够较好地保留图像的边界信息。由于BEMD算法可将图像分解为细节图像及余项图像,其所分解的细节图像包括DWI图像的边界信息以及主噪声,而余项图像则描述图像的趋势信息。因此,提出一种将二维经验模态分解算法与改进的维纳滤波器相结合的降噪算法,并将该算法应用于DWI图像的降噪中。通过实验,将所提出的算法与其他算法应用于DWI图像的降噪处理,并通过对结果的分析比较证明所提出的算法能够更有效地对DWI图像进行降噪处理。     

数字皮影动画脚本的研究与应用

为简化数字皮影动画的制作流程,提高数字皮影作品的交互性,降低数字皮影实时表演的难度,对皮影常见角色及动作进行了抽象、分类,提出一种针对特定类型的通用的骨骼模型。基于反向动力学,设计了一套描述皮影骨骼动画的脚本。该脚本通过对骨骼位置与旋转属性的控制定义皮影角色的基本动作,并通过对基本动作的叠加实现复杂动作。脚本对动作的定义基于骨骼模型而非具体角色,具有通用性。最后通过实验证明了脚本的有效性。     

螺杆挤压机混合能力的分析

分析螺槽中熔体的流动状态 ,运用加权平均总应变来评价螺杆挤压机的混合能力 ,进而提出提高混合能力的途径。对于单螺纹螺杆 ,稍长的计量段 ,稍浅的计量段螺槽及较高的螺杆转速 ,有利于混合作用 ;开发新型螺杆 ,如犁钉分流混合螺杆 ;研制高效混合型螺杆挤压机 ,如球窝型高效混合型螺杆挤压机     

Grafting poly(t‐butyl acrylate) to poly(allylamine) by inverse microemulsion radical polymerization: From comb‐polymer to amphiphilic shell crosslinked polymer nanocapsule

An interfacial grafting radical polymerization method for amphiphilic comb copolymer and shell crosslinked polymer nanocapsules was reported. Macropolyradicals on a water soluble long chain polyamine were generated with hydrogen peroxide in the water phase and subsequent grafting radical polymerization of a vinylic monomer at the water/oil interface proceeded at 65°C. In the presence of a crosslinker, the resulting graft copolymer formed a defined core‐shell structure with hydrophilic aqueous core functionalized by the polyamine and a hydrophobic crosslinked polymer shell. The structure of the core‐shell material was characterized by NMR, FTIR, DSC, TGA, SEM, TEM, and the mechanism of the graft polymerization is proposed. © 2007 Wiley Periodicals, Inc. J Appl Polym Sci 104: 1905–1911, 2007     

Designing amino acid sequences to fold with good hydrophobic cores

We present two methods for designing amino acid sequences ofproteins that will fold to have good hydrophobic cores. Giventhe coordinates of the desired target protein or polymer structure,the methods generate sequences of hydrophobic (H) and polar(P) monomers that are intended to fold to these structures.One method designs hydrophobic inside, polar outside; the otherminimizes an energy function in a sequence evolution process.The sequences generated by these methods agree at the levelof 60–80% of the sequence positions in 20 proteins inthe Protein Data Bank. A major challenge in protein design isto create sequences that can fold uniquely, i.e. to a singleconformation rather than to many. While an earlier lattice-basedsequence evolution method was shown not to design unique folders,our method generates unique folders in lattice model tests.These methods may also be useful in designing other types offoldable polymer not based on amino acids