基于局域相关向量机回归模型的小尺度网络流量的非线性预测

基于非线性时间序列局域预测法与相关向量机回归模型, 本文提出了局域相关向量机预测方法, 并应用于预测实际的小尺度网路流量序列. 应用基于信息准则的局域预测法邻近点的选取方法来选取局域相关向量机回归模型的邻近点个数. 对比分析了局域相关向量机预测法、前馈神经网络模型与局域线性预测法对网络流量序列的预测性能, 其中前馈神经网络模型的参数采用粒子群优化算法来优化. 实验结果表明: 邻近点优化后的局域相关向量机回归模型能够有效地预测小尺度网络流量序列, 归一化均方误差很小; 局域相关向量机回归模型生成的时间序列具有与原网络流量时间序列相一致的概率分布; 局域相关向量机回归模型的预测精度好于前馈神经网络模型的与局域线性预测法的. 关键词： 小尺度网络流量 非线性时间序列预测方法 局域预测法 相关向量机回归模型     

脉冲激光沉积法制备高温压电薄膜0.20 BiInO$lt;sub$gt;3$lt;/sub$gt;-0.80PbTiO$lt;sub$gt;3$lt;/sub$gt;$lt;strong$gt;$lt;font color="#ff0000"$gt;$lt;font color="#ff0000"$gt;（已撤稿）$lt;/font$gt;$lt;/font$gt;$lt;/strong$gt;

采用脉冲激光沉积法制备了0.20BiInO₃-0.80PbTiO₃（20BI-PT）高温压电薄膜,并与0.15BiInO₃-0.85PbTiO₃（15BI-PT）样品进行了比较研究. X射线衍射谱显示,20BI-PT样品100峰出现了明显的劈裂,显示样品具有更高的四方对称性. FESEM图显示,20BI-PT样品中出现了部分111取向的三角形晶粒. 20BI-PT样品的铁电剩余极化（P_r）为～28 μC/cm²,矫顽场（E_c）为～120 kV/cm,相较15BI-PT样品,P_r略有增加,但同时E_c也有增加. 20BI-PT样品的横向压电系数（e_31,f）约为–4.7±0.6 C/m²,和15BI-PT相比几乎一样. 介电温度谱显示,20BI-PT 样品的居里温度比15BI-PT增加了约30 ℃,达590 ℃,且介电峰没有明显的频率依赖性. Rayleigh分析显示,20BI-PT样品中内在本征因素及可翻转畴对介电非线性的贡献和15BI-PT基本相同,但是外在因素的贡献没有15BI-PT的贡献大,这可能和20BI-PT样品中晶粒111相对取向率较高有关. 关键词： 薄膜 脉冲激光沉积 铁电 压电     

Upgrade of beamline 3W1B at Beijing Synchrotron Radiation Facility

To improve the performance of Beamline 3W1B at the Beijing Synchrotron Radiation Facility for the soft X-ray magnetic linear dichroism research at transition metals L_{2, 3} edges, a new monochromator was designed and built to replace the original one. After the assemblage, alignment and adjustment of the monochromator system, the first commissioning results were obtained. The photon energy range is from 50 to 1000 eV with spectral resolutions of 1600 at 250 eV and 1000 at 870 eV. The photon flux is of the order of 10⁸-10⁹ photons/s/200 mA/0.1%BW. In the electron's orbital plane the linear polarization degree of the light is higher than 99% at 704 eV. The beamline has satisfied the basic experimental requirements.     

黏滞等离子体中双撕裂模不稳定性的数值模拟研究

本文采用磁流体力学模型,数值研究了平板位形下双撕裂模线性增长率关于等离子体电阻η和黏滞v的定标关系.结果表明,对于有理面间距较大的情况,线性增长率关于电阻和黏滞的指数定标率随着黏滞的增加逐渐由γ∝ η3/5v0的定标变化到γ∝η5/6v-1/6的定标;而对于有理面间距较小的情况,其指数定标率随着黏性的增加从γ∝η 1/3 v0的定标逐渐变化到γ∝η2/3v-1/3的定标.本文还给出了初始阶段对称的双撕裂模的非线性演化,发现在非线性阶段对称的双撕裂模将转化为反对称的双撕裂模,并解释了相应的物理机理.     

基于多重散射强度和偏振特征的舰船尾流气泡激光探测方法

研究多重散射效应对舰船尾流气泡群光散射强度和偏振特征的影响是舰船光尾流探测以及新型光自导鱼雷研究的基础. 基于矢量Monte Carlo方法建立了舰船尾流气泡群激光后向探测仿真模型, 重点研究了尾流气泡群的多重散射机理,分析了多重散射效应、尾流气泡群密度对回波信号强度和偏振特征的影响规律. 基于粒子碰撞重要性抽样的基本思想, 在传统能量接收方法的基础上, 提出了回波光子偏振贡献接收方法和回波信号偏振信息统计方法, 解决了小视场系统光子返回概率低无法形成回波能量的难题. 构建了模拟尾流气泡群激光散射强度和偏振探测实验平台, 从实验的角度验证了模拟结果的准确性. 实验和模拟结果的一致性表明, 利用回波强度、偏振信息可表征气泡群距离、密度信息, 从而可对舰船尾流特别是低密度尾流进行高精度的探测和辨识. 关键词： Monte Carlo 偏振 多重散射 气泡     

Efficiency and droop improvement in a blue InGaN-based light emitting diode with a p-InGaN layer inserted in the GaN barriers

The advantages of a blue InGaN-based light-emitting diode with a p-InGaN layer inserted in the GaN barriers is studied. The carrier concentration in the quantum well, radiative recombination rate in the active region, output power, and internal quantum efficiency are investigated. The simulation results show that the InGaN-based light-emitting diode with a p-InGaN layer inserted in the barriers has better performance over its conventional counterpart and the light emitting diode with p-GaN inserted in the barriers. The improvement is due to enhanced Mg acceptor activation and enhanced hole injection into the quantum wells.     

Progress in superconducting qubits from the perspective of coherence and readout

Superconducting qubits are Josephson junction-based circuits that exhibit macroscopic quantum behavior and can be manipulated as artificial atoms.Benefiting from the well-developed technology of microfabrication and microwave engineering,superconducting qubits have great advantages in design flexibility,controllability,and scalability.Over the past decade,there has been rapid progress in the field,which greatly improved our understanding of qubit decoherence and circuit optimization.The single-qubit coherence time has been steadily raised to the order of 10 to 100μs,allowing for the demonstration of high-fidelity gate operations and measurement-based feedback control.Here we review recent progress in the coherence and readout of superconducting qubits.     

芬布芬-铜（Ⅱ）-牛血清白蛋白相互作用的光谱法研究

模拟生理条件下,应用荧光光谱法研究了芬布芬对牛血清白蛋白,铜（Ⅱ）对牛血清白蛋白以及铜（Ⅱ）对芬布芬和牛血清白蛋白荧光光谱特性的影响.实验表明：铜（Ⅱ）和芬布芬均可使牛血清白蛋白的荧光强度发生静态猝灭,并且在铜（Ⅱ）存在下,芬布芬对牛血清白蛋白的荧光猝灭作用显著增强.根据荧光猝灭双倒数图计算芬布芬和牛血清白蛋白的结合常数为8.44×104,结合位点数为0.97.荧光猝灭双倒数图计算的结果表明,芬布芬和牛血清白蛋白之间的结合常数和结合位点数均随铜（Ⅱ）浓度的增大而增大.很据三者结合反应的研究,进一步探讨了芬布芬、铜（Ⅱ）在生物体内与蛋白质相互作用的机理.