The dynamics of triple-well trapped Bose-Einstein condensates with atoms feeding and loss effects

In this paper, we consider the macroscopic quantum tunnelling and self-trapping phenomena of Bose-Einstein condensates （BECs） with three-body recombination losses and atoms feeding from thermal cloud in triple-well potential. Using the three-mode approximation, three coupled Gross-Pitaevskii equations （GPEs）, which describe the dynamics of the system, are obtained. The corresponding numerical results reveal some interesting characteristics of BECs for different scattering lengths. The self-trapping and quantum tunnelling both are found in zero-phase and ：r-phase modes. Furthermore, we observe the quantum beating phenomenon and the resonance character during the self-trapping and quantum tunnelling. It is also shown that the initial phase has a significant effect on the dynamics of the system.     

弱Hopf量子Yang-Baxter H-模基本定理(英文)

本文引入弱Hopf量子Yang-Baxter模概念.利用弱Hopf模基本定理的方法,获得了弱Hopf量子Yang-Baxter模基本定理,进一步还得到了相关Hopf模基本定理.     

The stability of Bose--Einstein condensates in a two-dimensional shallow trap

The stability of Bose--Einstein condensates (BECs) loaded into a two-dimensional shallow harmonic potential well is studied. By using the variational method, the ground state properties for interacting BECs in the shallow trap are discussed. It is shown that the possible stable bound state can exist. The depth of the shallow well plays an important role in stabilizing the BECs. The stability of BECs in the shallow trap with the periodic modulating of atom interaction by using the Feshbach resonance is also discussed. The results show that the collapse and diffusion of BECs in a shallow trap can be controlled by the temporal modulation of the scattering length.     

中学物理静电演示实验研究

分析了中学物理教学法静电演示实验所存在的困难,建议从实验的材料和趣味性来提高静电演示实验教学效果,并介绍一些低成本高教育价值静电实验,以期望对中学物理教学法演示实验课程的开展有所帮助.     

一种基于重置的变结构前馈神经网络

基于GaussNewton法的前馈神经网络虽然可以达到局部二阶收敛速度．但网络结构中如果结点个数过多，会造成过模拟；网络结点过少。又会导致不收敛。为了优化神经网络结构，尝试引入重置算法(Early Restart Algofithm)，并将其应用于Gauss Newton前馈神经网络．提出基于重置的Gauss Newton变结构前馈神经网络。对比实验表明，重置算法的引入有效地解决神经网络的结构优化问题，优化后的神经网络具有良好的收敛性与稳定性。     

双边弱smash积的Maschke型定理

本文主要给出半单弱Hopf代数上双边弱Smash积的Maschke型定理.     

ZnO薄膜的自组织设计及形貌控制

采用单源化学气相沉积(SSCVD)法,在Si(100)基片上通过改变前驱反应体与基片的入射角度,获得了可控柱状取向的ZnO薄膜.研究发现,入射角度的改变可使沉积薄膜中的柱状结构的生成方向倾斜.但X射线衍射(XRD)分析表明:ZnO薄膜的c轴(002)取向与入射角无关,且不沿ZnO柱状结构的生长方向取向.由于ZnO的(002)面为其表面自由能最低且原子密度高的晶面,ZnO薄膜的生长更易于在垂直于基片表面的方向c轴取向生长.