超细聚苯乙烯(PS)树脂的制备工艺研究

采用O/W型苯乙烯微乳液体系,研究了乳化剂OP-10、助乳化剂DBS、交联剂DVB等对聚苯乙烯(PS)粒径分布的影响,基于正交试验,找出了量佳微乳液体系配方。采用激光粒度仪与透射电镜(TEM)对纳米PS粒子的粒径进行了测量研究,结果表明:交联PS粒子单分散性好,大小较均匀。     

Planar and axially symmetric slow shearing granular flows: a comparative evaluation of assembly evolution incorporating wavelet transforms

We demonstrate the enhanced capacities of our analysis toolset for three-dimensional data. In particular, we provide supporting evidence for some of the conclusions reached in our previous studies of two-dimensional avalanching heaps. Segregation by self-diffusion is shown to take place in three-dimensional assemblies and self-diffusion velocities are shown to be of a comparable order of magnitude to those found in plane-strain situations. The effect on assembly evolution by discrete avalanching of the availability of a third dimension for translation is investigated. The discrete wavelet transform is again shown to be a useful component of the toolset in coupling process variables in the context of the discrete defining events associated with assembly evolution. In particular, we move towards the determination of time constants by correlating wavelet coefficients with a time shift.     

中国建筑涂料工业的现状及当前形势下的发展对策

本文详细分析了中国建筑涂料工业的现状和存在问题，并结合当前形势，提出了中国建筑涂料工业的发展对策。     

一种结合主题模型的推荐算法

针对传统协同过滤推荐算法存在的冷启动、数据稀疏以及相似度度量的准确性问题,基于LDA主题模型对文本隐式主题挖掘的有效性和KL散度在主题分布相似性度量的准确性,提出了结合LDA主题模型的矩阵分解推荐算法。首先,利用改进的LDA算法输出项目-主题分布,并用困惑度作为主题数设置的修正函数;然后分别基于余弦相似度和KL散度计算得到项目相似度矩阵,将得到的相似度矩阵结合原评分训练集输出预评分,再将预评分填充到训练集;最后将训练集输入ALS矩阵分解算法得到推荐结果。通过MovieLens数据集的实验结果表明,该算法在不同隐式参数设定下均能得到比ALS推荐算法以及更小的预测误差,并且最优预测误差小于传统推荐算法。该实验说明了通过集成LDA主题模型的ALS算法效果要优于其他推荐算法。     

凝胶过滤色谱法研究酶解过程中木聚糖分子结构的变化

采用凝胶过滤色谱法研究了木聚糖在酶解过程中分子量分布的变化。结果表明,木聚糖酶对高分子量组分的木聚糖的降解能力较差,而主要降解分子量介于15000～3000之间组分的木聚糖,使得这部分木聚糖变为分子量更低的木聚糖组分。随着酶解时间的延长,酶解得率上升,酶解产物中低聚木糖含量增多,未被酶解的木聚糖的平均聚合度上升,当酶解时间超过4h时,这种变化趋势缓慢。随着酶解轮次的增加,酶解产物中可溶性木聚糖的含量越来越少,木聚糖被降解的难度越来越大。在实际生产中,用来作为低聚木糖生产的木聚糖其聚合度应在20～100之间,酶解时间以4h为宜,重复利用未水解木聚糖的轮次以2轮为宜。     

基于LDA和word2vec的英文作文跑题检测

针对目前国内的英语作文辅助批阅系统缺少准确而高效的跑题检测算法的问题,提出了一种结合LDA和Word2vec的跑题检测算法。该算法利用LDA模型对文档建模并通过Word2vec对文档训练,利用得到的文档主题和词语之间的语义关系,对文档中各主题及其特征词计算概率加权和,最终通过设定合理阈值筛选出跑题作文。实验中通过改变文档的主题数而得到的不同F值,确定了最佳主题数。实验结果表明所提出的方法比基于向量空间模型的方法更具有效性,可以检测到更多的跑题作文并且准确率较高,F值达到89%以上,实现了作文跑题检测的智能化处理,可以有效地应用在英语作文教学中。     

Nonparametric Bayesian Image Segmentation

Image segmentation algorithms partition the set of pixels of an image into a specific number of different, spatially homogeneous groups. We propose a nonparametric Bayesian model for histogram clustering which automatically determines the number of segments when spatial smoothness constraints on the class assignments are enforced by a Markov Random Field. A Dirichlet process prior controls the level of resolution which corresponds to the number of clusters in data with a unique cluster structure. The resulting posterior is efficiently sampled by a variant of a conjugate-case sampling algorithm for Dirichlet process mixture models. Experimental results are provided for real-world gray value images, synthetic aperture radar images and magnetic resonance imaging data.     

Multi-rate model predictive control of particle size distribution in a semibatch emulsion copolymerization reactor

A multi-rate model predictive controller algorithm is presented for the in-batch closed-loop control of the full particle size distribution (PSD) in a semibatch emulsion copolymerization system. The lack of frequent measurements of the PSD and the measurement delay of these measurements are addressed through the use of frequent density measurements from which the current conditions of the system are estimated. The high dimensionality of the discretized full PSD is reduced by the use of model order reduction based on principal component analysis. This method effectively reduces the size of the problem while preserving the main characteristics of the population balance system. Disturbances that perturb the surfactant and monomer amounts inside the semibatch vinyl acetate–butyl acrylate reactor are considered to demonstrate the performance of the proposed control algorithm.