Molecular association complex of urea with polyethylene: 3. Structural studies of the complex

The crystal structure of urea-polyethylene complex in the hexagonal form, obtained by one of our preparation methods, is analysed by X-ray power diffractometry. In this analysis, the fixed molecular parameters of urea are used, and three models with respect to rotational disorder of the guest polyethylene molecule are assumed. It is shown that the host lattice structure constructed by urea molecules is essentially the same as that of urea-n-paraffin complex. Effects of the disorder structure of the guest on X-ray scattering of the complex are presented. However, there is a difficulty in identifying the disorder structure by X-ray analysis. With the aid of a potential energy calculation between the hexagonal urea tunnel and the guest molecule, a rotational disorder model is preferred.     

Configurational statistics of copolymers prepared by complex-radical polymerization

Radical copolymerization is considered theoretically taking into account the effect of complexing on the configurational statistics of macromolecules formed. Akinetic model has been developed that considers, apart from the ordinary addition of single monomer units to a propagating chain, the possible addition of monomer unit pairs along with a complexing agent in the form of a ternary complex. Within the framework of this model, the problem of calculating the probabilities of formation of any sequences of monomer units (taking into account their microtacticity) in a macromolecule has been rigorously solved as well as the problem of finding the composition distribution of the copolymer formed. It has been shown that this distribution is described by a conventional Gauss law and the appropriate parameters are given. Possible generalizations of the suggested approach are indicated.     

Multi-label core vector machine with a zero label

Multi-label core vector machine (Rank-CVM) is an efficient and effective algorithm for multi-label classification. But there still exist two aspects to be improved: reducing training and testing computational costs further, and detecting relevant labels effectively. In this paper, we extend Rank-CVM via adding a zero label to construct its variant with a zero label, i.e., Rank-CVMz, which is formulated as the same quadratic programming form with a unit simplex constraint and non-negative ones as Rank-CVM, and then is solved by Frank–Wolfe method efficiently. Attractively, our Rank-CVMz has fewer variables to be solved than Rank-CVM, which speeds up training procedure dramatically. Further, the relevant labels are effectively detected by the zero label. Experimental results on 12 benchmark data sets demonstrate that our method achieves a competitive performance, compared with six existing multi-label algorithms according to six indicative instance-based measures. Moreover, on the average, our Rank-CVMz runs 83 times faster and has slightly fewer support vectors than its origin Rank-CVM.     

Crystalline complex between poly(γ-methyl l-glutamate) and dimethyl phthalate

The crystalline complex between poly(γ-methyl l-glutamate) (PMLG) and dimethyl phthalate (DMP) has been formed in films cast from a solution in dichloroethane. It has the stoichiometry of 1 mol of DMP to three or four residues of PMLG and shows two definite characteristics in X-ray diffraction patterns; one is the large hexagonal unit cell with the edge of around 28 Å and another the ‘extra’ 5.07 Å meridional reflection which can not be interpreted by a PMLG α-helical conformation. The structural examination for the films with various DMP contents is carried out by X-ray, viscoelastic, and d.s.c. measurements and the following structure is proposed for the crystalline complex. Four PMLG are associated to form a group which is hexagonally packed and DMP molecules, located in the gaps between groups, form a specific favourable helical structure along PMLG chains in which the van der Waals stacking of benzene rings of DMP is significant.     

基于支持向量回归的动力电池参数辨识

根据铅酸蓄电池在放电过程中内部电化学反应导致外部电特性变化的特点,提出一种基于支持向量机原理的电解液密度辨识模型。利用支持向量机理论非线性回归的特性,简化测量电解液密度的过程,在恶劣环境下检测动力电池的电解液密度更显其优越性。预测实验表明,采用改进的交叉验证预测模型具有泛化能力强、稳定性好的特点,并且在小样本的条件下能达到预期的辨识精度。     

基于概率支持向量回归的产品设计时间预测模型

为使产品设计时间预测既克服小样本、异方差噪声问题,又提供除预测值以外的其他有用信息,建立概率支持向量回归(PSVR)模型。首先,在异方差回归模型基础上设计概率约束条件,以使预测值以较大概率位于真实值的某邻域,结合具有参数不敏感损失函数的支持向量回归确定优化目标,提出PSVR。然后,将最大完工时间知识嵌入进PSVR的约束条件,用以确定真实值邻域的宽度,将交叉验证与遗传算法相结合以确定PSVR的相关参数。最后,以注塑模具设计的实例进行分析,结果表明基于PSVR的时间预测方法可同时提供有效的预测值和预测区间。     

PSO-SVM模型的构建与应用

为了使支持向量机(SVM)获得更好的分类效果,针对人为选择参数的随机性,提出了利用粒子群算法(PSO)进行参数自动选取的优化方法,构建了PSO-SVM模型.在个人信用评估中,通过对粒子适应度函数的设置来控制造成较大损失的第二类误判,应用结果表明:模型在训练和测试样本中的分类精度可以达到95%,第二类误判率分别仅为0.78%和2.02%.利用PSO对SVM中的参数进行优化,可以避免人为选择的随机性,并且在解决分类问题中表现出较好的稳健性.     

基于多特征融合的运动物体识别方法研究

为了实现运动目标的实时检测,本文基于多特征融合,提出了一种运动物体分级识别方法.首先通过对运动目标的分割,分析得到各个物体的形状矩特征、面积大小及运动速度.对上述特征,采用支持向量机方法对其矩特征进行快速的一级分类,然后对一级分类结果及各物体的面积、速度等信源根据其对分类的重要程度,采用模糊积分技术进行多特征融合分类.实验表明,本方法具有较强的学习能力及较高的识别精度,可有效地应用于道路监控等系统.     

汉语语音文档检索中后验概率的索引方法

基于音节Lattice形式的语音识别结果来实现汉语语音文档检索,不但可以成功规避词表外词问题,而且Lattice这种多候选形式也能有效补偿识别错误对检索性能的影响.在基于音节Lattice的汉语语音文档检索研究中,针对已有索引方法的不足,提出了一种基于后验概率的索引方法,对向量空间模型进行改进,以音节和K步邻接音节对作为索引项,以它们在语音文档中的后验概率值作为索引项权重.检索实验表明,文中的方法更适用于基于音节Lattice的语音文档检索任务,各项改进都达到了预期效果.     

海洋平台灰预测和支持向量机的逆控制

导管架式海洋平台在随机波浪等外载荷作用下极易产生有害振动,且其动力响应具有极强的非线性和时变性,采用被动控制方法和基于精确数学模型的主动控制方法控制海洋平台的有害振动很难达到理想的控制效果.结合灰预测和支持向量机的优点,提出了基于灰预测和支持向量机的自适应逆控制方法,并将其与动态刚度阵法相结合应用于导管架式海洋平台的振动主动控制中.数值算例分析表明,该控制方法可有效地控制波浪和风载荷作用下引起导管架式平台的有害振动,并能解决由于控制信号传输等原因引起的时滞问题.