玻尔兹曼极限下非增长复杂网络中的凝聚（英文）

讨论了复杂网络中几种不同的凝聚现象.分析了具有固定结点数和固定连边数的非增长适应度模型中的临界现象.解析研究的结果表明这一模型展示出了凝聚相变的现象.同时用数值方法模拟了该模型在玻尔兹曼极限下的临界行为,发现其非凝聚相与增长适应度模型中的非凝聚相有显著差异,而对应的临界温度和度分布也被确定下来.     

自适应粒子群神经网络交通流预测模型

针对传统神经网络预测模型预测结果准确性低且存在大量无效迭代的问题,提出了自适应权重粒子群神经网络交通流预测(PSOA-NN)模型。首先根据待预测点的上下游观测点数和历史数据,随机初始化若干组模型参数并计算每组参数对应粒子的适应度;然后采用改进的sigmoid函数替代原有模型中的固定惯性权重,并根据其中适应度变好的粒子更新粒子速度和位置,一直迭代到粒子适应度小于预设值为止;最后将满足条件粒子对应的模型参数应用到神经网络模型,根据实时交通流数据预测出15min后的数据。仿真结果表明,使用PSOA-NN模型,可使得在同等预测误差范围内收敛速度提升0.6~1.7倍。     

凝聚体中的耗散现象若干应用

本书从非平衡统计物理学介绍了耗散现象的模型化和品质鉴定，论述了凝聚物质物理学、材料科学、化学物理学等方面。     

一种非共轴应变硬化本构模型数值应用及离心机试验验证

在土体的剪切变形过程中,当主应力方向产生旋转时,主应变增量方向与主应力方向之间存在显著的非共轴现象.同时,机动摩擦角、膨胀角随着累积塑性偏应变的增长而增加,土体具有应变硬化的特点.传统的弹塑性本构模型不能够反映上述现象对地基承载力特性的影响.为了能够对地基承载力问题进行合理的分析,建立了一种非共轴应变硬化模型,并将该模型运用到有限元计算中.通过与三轴试验和离心机模型试验结果进行对比,对该模型在数值应用中的合理性进行了验证.研究结果表明,该模型能够对不同围压下的应力-应变关系进行预测.对浅基础承载力问题进行研究时,非共轴应变硬化模型的计算结果比传统弹塑性本构模型更加接近于离心机试验结果,验证了该模型的数值应用合理性.     

用非连续介质力学模型研究混凝土的Ⅰ型断裂

为进行连续介质到非连续介质转化的数值模拟,实现结构破坏过程的仿真分析,将固定和旋转的弥散裂缝模型与变形体离散元方法结合应用于混凝土、岩石等准脆性材料的受拉开裂过程分析。通过对比三点弯梁试验的数值模拟结果与试验结果,验证了模型能够定量预测裂缝的发生及开裂过程中的变形,获得了合理的结构荷载位移响应曲线,同时旋转裂缝模型同固定裂缝模型相比可以一定程度上避免剪切锁死现象。方法为连续非连续介质统一力学模型的进一步研究奠定了一定的理论基础。     

基于遗传-神经网络的无粘结部分预应力高强混凝土梁延性预测模型

尝试利用遗传-神经网络模型,对无粘结部分预应力高强混凝土梁的延性进行预报.选用23个试件为学习样本,3个试件为测试样本.模型得到了很好的预测结果.模型中使用了最优保存策略和实数编码.同时,为了防止＂早熟＂现象的发生,提出了一种新的改变适应度值的方法.     

基于空间面板数据模型的区域经济增长研究

为分析我国省际及各时段经济增长的影响因素,在空间面板数据模型基础上加入时间和空间固定效应,建立无固定效应模型、空间（地区）固定效应模型、时间固定效应模型和时空（时间空间）混合固定效应模型等4个不同的模型,来拟合分析2001～2010年中国31个省市的经济数据.发现,时空混合效应面板数据模型的拟合度最高,中国各地区的经济增长的时空间依赖关系是显著的,全国各地区的经济增长不仅具有地区差异性,还呈现出阶段性特征.     

自适应遗传算法

自适应遗传算法采用自适应的适应度函数、交叉概率及变异概率代替固定的适应度函数、交叉概率及变异概率,与基本遗传算法相比,结果证明改进的遗传算法显著提高了收敛性能,并且具有很强的自适应能力.     

Banach空间二阶边值问题解的存在性

讨论了Banach空间中非线性二阶Dirichlet边值问题解的存在性.在非线性项满足较弱的非紧性测度条件及线性增长条件下,应用凝聚映射的拓扑度理论获得了该问题解的存在性.这个线性增长条件是保证解存在的最优条件.     

均匀增长无标度网络的等价模型

分析了无标度网络的BA模型和与BA等价的随机连接到边模型.在给出边的权重和节点适应度定义的基础上,提出了无标度网络的边加权网络模型和节点适应度模型.通过解析的方法证明了它们与BA模型的等价性.计算机模拟的结果也说明了它们的等价性,并给出了4个等价模型之间的比较.     

防止重掺砷CZ单晶硅组分过冷的探讨

在探讨组分过冷数学模型的基础上,针对重掺砷CZ单晶硅的生长,理论计算了防止组分过冷时固液界面处晶体温度梯度G_S的临界值为51.32~33.10 K/cm.以此为依据,设计了具有较大温度梯度的18寸(60 cm)晶体生长热场,以数值模拟的方法,给出了固液界面处晶体的温度梯度G_S的模拟值为54.68~38.14 K/cm.在晶体等径生长的各个阶段,固液界面处晶体的温度梯度G_S的模拟值均在防止组分过冷的临界值之上,可以有效避免晶体生长过程中组分过冷的发生,并利用实际晶体生长试验的结果验证了以上分析的有效性.     

Monte Carlo simulations for two-dimensional Ising system far from equilibrium

With a dynamic Monte Carlo simulation, which is free of critical slowing down, the critical behavior of two-dimensional Ising model at non-equilibrium states is investigated. We focus on the two-time autocorrelation function A(t, t' ) to identify the dynamic exponent z in two different evolution stages, quenched from a high temperature state and a completely ordered state to the critical state at TC, re-spectively. By using the heat-bath algorithm, the dynamic critical exponent is estimated as z ≈ 2.16. As a result, the universality of the scaling behavior is verified numerically, and the exponent λC is deter-mined complete by λC = β /ν for the evolutions from an initial completely ordered state.     

A mean-field Bak-Sneppen model with varying interaction strength

We propose a mean-field Bak-Sneppen (MFBS) model with varying interaction strength. The interaction strength, here denoted by α, specifies the degree of interaction, and varies smoothly between 0 for no interaction and 1 for full interaction (restoring the original BS model). Our simulations of the MFBS model reveal some interesting features. When α is non-zero, the MFBS model can evolve to a self-organized critical (SOC) state. The critical exponent of the avalanche size distribution, α, is insensitive to ...     

In-situ Mg77Cu12Zn5Y6 bulk metallic glass matrix composite with extraordinary plasticity

Mg base BMG alloys have attracted a great deal of attention for its low density and relative low cost. To date, BMG alloys with millimeter scale have been fab- ricated by using conventional copper cast and die cast method in Mg-Ni-Ce[1], Mg-Ni-La[2], Mg-C…     

High undercooling and rapid dendritic growth of Cu-Sb alloy in drop tube

Droplets of Cu-20%Sb hypoeutectic alloy has been rapidly solidified in drop tube within the containerless condition. With the decrease of droplet diameter, undercooling increases and the microstructures of primary copper dendrite refines. Undercooling up to 207 K (0.17 T _L) is obtained in experiment. Theoretic analysis indicated that because of the broad temperature range of solidification, the rapid growth of primary copper dendrite is controlled by the solutal diffusion. Judging from the calculation of T₀ curve in the phase diagram, it is shown that the critical undercooling of segregationless solidification is δT ₀ = 474 K. At the maximum undercooling of 207 K, the growth velocity of primary copper phase exceeds to 37 mm/s, and the distinct solute trapping occurs.     

Harmonic hexagonal superlattice pattern in a dielectric barrier discharge at atmospheric pressure

We report a harmonic hexagonal superlattice pattern in a dielectric barrier discharge in air/argon mixture at atmospheric pressure. The bifurcation scenario of harmonic hexagonal superlattice pattern with the applied voltage increasing is given. The phase diagram of the pattern types as a function of the applied voltage and the air-concentration is obtained. The hysteresis of pattern transitions at the upward and downward stage of the applied voltage is observed. The correlation measurements indicate that harmonic hexagonal superlattice pattern is an interleaving of two different transient sublattices. The spatial power spectrum demonstrates that harmonic hexagonal superlattice pattern has two separate wave vectors. Both small wave vector qh and big wave vector Kh, belong to the harmonic mode, and they obey a triad resonant interaction qh^1 ＋ qh^2 = Kh.     

Transitions of amorphouscrystalline-amorphous in bulk metallic glass under HP and HT

In-situ SR-XRD measurements revealed that the crystallization process in Zr₄₁.2Ti₁₃.8Cu_12.5Ni₁₀Be_22.5 bulk metallic glass is significantly different from that in traditional glasses. Subsequent heating at 10 GPa converts the sample from amorphous phase into the metastable fcc phase and then leads to the fcc phase back to the amorphous phase, indicating that there exists ‘reversible’ phase transition phenomena in the material under high pressure and high temperature.     

QSPR between Physical-Chemical Properties and Molecule Parameters of Alkanes

A set of molecule parameters, namely, N, N′, p, q, n, were used to express the structures of alkanes. A correlative model was established between certain physical-chemical properties and molecular parameters of alkanes by regression method. Eight physical-chemical properties, such as evaporation heat （△vHm^20）, density（D^20 ）, capacity （C^20）, surface tension （δ^20）, boiling point （Tb）, critical temperature（Tc）, critical pressure（Pc） and critical volume（Vc）, of fifty-six C3-C16 alkanes were calculated directly from the model in this paper. The calculated values are in good accordance with the literature ones reported for alkanes, and the correlation coefficients （R） equal or exceed 0.99 . The research results indicate that the principle of the method is simple and clear, the method is practical, the correlativity is excellent, and the predicted data are credible.     

Experimental studies on interaction of water mist with class K fires

Interaction of water mist with cooking oil fires is studied experimentally and theoretically. A LDV/APV system is used to measure the velocity and diameter of water mist at different pressures in the experiments,and the effect of water mist velocity and diameter on fire extinguishment efficiency is investi-gated. The experimental results show that water mist has excellent surface cooling effect; it can control and extinguish cooking oil fires quickly without re-ignition. The critical temperature (Tfo) is calculated by energy balance equation,and the fire plume momentum is calculated and compared with that of water mist in order to determine the critical velocity (νwy) of fire extinguishment. This paper provides references for cooking oil fires extinguishment with water mist.     

Core-shell microstructure formed in the ternary Fe-Co-Cu peritectic alloy droplet

The metastable liquid phase separation occurs in the ternary Cu₅₀Fe_37.5Co_12.5 peritectic alloy droplets during free fall. The separated alloy melt rapidly solidifies and evolves core-shell microstructure composed of L₁(Cu) and L₂(Fe,Co) phases. Based on the determination of the phase transition temperature, the core-shell microstructure evolution, the interfacial energy, the temperature gradient and the Marangoni migration are analyzed. The interfacial energy of the separated liquid phase increases with the decrease of the temperature. The temperature gradient changes from large to small along the radius direction from inside to outside in the alloy droplet. The Marangoni force (F _M) acting on the micro-droplet of L₂(Fe,Co) phase increases with the increase of the size of the L₂(Fe,Co) phase, and decreases with the increase of undercooling. Driven by F _M, the micro-droplet of L₂(Fe,Co) phase migrates from outside to inside in the alloy droplet, collides and coagulates each other during migration, and then forms different types of core-shell microstructures. Supported by the National Natural Science Foundation of China (Grant Nos. 50121101, 50395105) and NPU Youth Scientific and Technological Innovation Foundation (Grant No. W016223)