改进型单晶炉提拉系统偏心平衡研究

随着光伏行业的快速发展, 对硅单晶的品质和长晶装备的稳定性的要求也不断提高。直拉法是生产硅单晶的主要方法,通过提高单晶炉副室的高度以扩大单晶硅的生产规模。由于副室高度的大幅增加,且单晶炉提拉头质心相对于旋转轴心有一定距离,对单晶炉整体稳定性有较大影响,从而降低了单晶硅的生产质量。针对此问题,对单晶炉建立可靠的力学分析模型,采用数值仿真方法,对单晶炉整体进行动力学响应分析,计算得到副室高度增加后的单晶炉工作时中钨丝绳下端晶棒的运动规律以及最大摆动幅度,为改进设计提供依据。数值仿真分析表明提高单晶炉副室高度后,提拉头较大的质心偏心是单晶炉提拉系统发生摆动的主要原因。在此基础上提出在提拉头上添加质心调节装置,通过控制系统调节可保证提拉头质心位置在旋转轴线上以降低提拉系统的摆动。     

Modes conversion due to plasmons induced transparency

We investigate simulated and theoretically the optical properties of the metamaterial, composed of two bigger split ring resonators and one smaller split ring resonator in unit cell. We observe the magic phenomena that one mode B is inhibited from strong to weak then disappears while another mode A appears and becomes stronger and stronger as asymmetric degree increases. The results show the mode A originates from the destructive interference between the dipole mode and the quadrapole mode, and its strength is proportional to the cross coupling coefficient of near-field. The disappearance of the mode B is due to the competition between the mode B and the mode A, and the variation of strength is proportional to the frequency shift of the dark mode. That is, with asymmetric degree increasing, the mode B converts into the mode A. These phenomena are explained very well by the temporal coupled-mode theory. Our metamaterial provides a kind of new design for understanding the interaction between light and matter.     

Temporal logic control for stochastic linear systems using abstraction refinement of probabilistic games

We consider the problem of computing the set of initial states of a dynamical system such that there exists a control strategy to ensure that the trajectories satisfy a temporal logic specification with probability 1 (almost-surely). We focus on discrete-time, stochastic linear dynamics and specifications given as formulas of the Generalized Reactivity(1) fragment of Linear Temporal Logic over linear predicates in the states of the system. We propose a solution based on iterative abstraction-refinement, and turn-based 2-player probabilistic games. While the theoretical guarantee of our algorithm after any finite number of iterations is only a partial solution, we show that if our algorithm terminates, then the result is the set of all satisfying initial states. Moreover, for any (partial) solution our algorithm synthesizes witness control strategies to ensure almost-sure satisfaction of the temporal logic specification. While the proposed algorithm guarantees progress and soundness in every iteration, it is computationally demanding. We offer an alternative, more efficient solution for the reachability properties that decomposes the problem into a series of smaller problems of the same type. All algorithms are demonstrated on an illustrative case study.     

基于局部均值的K-近质心近邻光谱分类

天体光谱包含着许多重要的关于天体的物理和化学信息,如天体表面的有效温度、重力加速度以及化学丰度等,天体光谱的处理和分析对天文研究具有重要的科学意义。一些大型巡天计划的实施(如SDSS,LAMOST等)使我们获得了海量的天文光谱数据,因此天文光谱数据的自动分类成为重要的科学研究课题,然而面对如此海量的光谱数据,一些传统的光谱自动分类方法已经不适用,迫切需要寻找高效率的光谱自动分类技术。研究了基于局部均值的K-近质心近邻(local mean-based K-nearest centroid neighbor,LMKNCN)算法在恒星(Star)、星系(Galaxy)和类星体(Quasar,QSO)的光谱分类中的应用。LMKNCN算法的基本思想是根据近质心近邻原则,从每一类训练样本集中为待测样本点选取k个近质心近邻点,然后根据每一类中所选取的k个近质心近邻点的均值点到待测样本点x的距离来判别x的所属类别。针对美国SDSS-DR8的天体光谱数据,对比了K-近邻、K-近质心近邻、LMKNCN三种算法在恒星、星系和类星体的光谱分类中所表现的性能,结果表明三种方法中,LMKNCN算法对这三种光谱的识别率高于其他两种算法或者与其相当,而且其平均分类正确率高于另外两种算法,特别是在类星体的识别率上表现的更好。表明了该算法对天文光谱大数据的快速处理和有效利用具有重要的意义。     

Temporal freedom in mathematical thought: A philosophical–empirical enquiry

This paper seeks to introduce and elaborate on the notion of ‘temporal freedom’ in the context of mathematical thinking.The intention is to merge a philosophical and empirical discussion on this topic, and findings will be presented from a qualitative study in which 29 high-school students were invited to explore the symmetry group of a square during clinical interviews. The data indicate how the interviewees – who were given the opportunity to freely explore an unseen mathematical system – repeatedly brought past and future together in a tangible present during their investigations. In doing so they were exceptionally successful in engaging with advanced abstract algebra, far beyond what might have been expected at the outset.By using empirical data to exemplify an otherwise highly philosophical discussion, I hope to draw explicit attention to the temporal aspect of the human mind in mathematical action which instructional methods frequently fail to recognise, value and utilise.     

Temps de cohérence d'un condensat de Bose–Einstein dans un gaz isolé harmoniquement piégé

We study the condensate phase dynamics in a low-temperature equilibrium gas of weakly interacting bosons, harmonically trapped and isolated from the environment. We find that at long times, much longer than the collision time between Bogoliubov quasi-particles, the variance of the phase accumulated by the condensate grows with a ballistic term quadratic in time and a diffusive term affine in time. We give the corresponding analytical expressions in the limit of a large system, in the collisionless regime and in the ergodic approximation for the quasi-particle motion. When properly rescaled, they are described by universal functions of the temperature divided by the Thomas–Fermi chemical potential. The same conclusion holds for the mode damping rates. Such universality class differs from the previously studied one of the homogeneous gas.     

Information filtering in evolving online networks

Recommender systems use the records of users' activities and profiles of both users and products to predict users' preferences in the future. Considerable works towards recommendation algorithms have been published to solve the problems such as accuracy, diversity, congestion, cold-start, novelty, coverage and so on. However, most of these research did not consider the temporal effects of the information included in the users' historical data. For example, the segmentation of the training set and test set was completely random, which was entirely different from the real scenario in recommender systems. More seriously, all the objects are treated as the same, regardless of the new, the popular or obsoleted products, so do the users. These data processing methods always lose useful information and mislead the understanding of the system's state. In this paper, we detailed analyzed the difference of the network structure between the traditional random division method and the temporal division method on two benchmark data sets, Netflix and MovieLens. Then three classical recommendation algorithms, Global Ranking method, Collaborative Filtering and Mass Diffusion method, were employed. The results show that all these algorithms became worse in all four key indicators, ranking score, precision, popularity and diversity, in the temporal scenario. Finally, we design a new recommendation algorithm based on both users' and objects' first appearance time in the system. Experimental results showed that the new algorithm can greatly improve the accuracy and other metrics.     

Trends of Atmospheric CCl 3 F,CCl 2 F 2 and CCl 2 FCClF 2 in Shanghai (China)

Atmospheric concentrations of CCl 3 F (CFC-11), CCl 2 F 2 (CFC-12), and CCl 2 FCClF 2 (CFC-113) in troposphere in Shanghai, China (31°1N) have been routinely monitored by using gas chromatography with electron capture detector from November 1997 to December 2000. The observation shows that there is a slowly declining trend of the three compounds. On the other hand, a comparison has been drawn between the concentrations observed and the ones calculated by 2-box model on the basis of the global statistical data of CFCs production and emission. Our observation agrees with the calculations. It shows that production and emission of chlorofluorocarbons (CFCs) in China are still at a low level and gradually decreasing.     

Control of coexisting attractors via temporal feedback

We propose a method for controlling coexisting attractors in multistable dynamical systems. In this method, the feedback for an initial duration of time can drive the system to the desired state. We have illustrated this method by considering temporal feedback in autonomous as well as non-autonomous dynamical systems. The experimental realisation of the proposed method is also presented.