求解考虑颗粒凝并的通用动力学方程的多重MonteCarlo算法

Monte Carlo(MC)方法被广泛用于通用动力学方程的求解,然而普通MC方法的计算代价较高而计算精度不稳定．提出一种新的多重Monte Carlo(MMC)算法来求解GDE,该算法同时具有基于时间驱动MC方法、常数目法和常体积法的特点．首先详细介绍了该算法,包括加权虚拟颗粒的引入,MMC算法的计算流程,时间步长的设置,颗粒是否发生凝并事件的判断,凝并伙伴的寻找,凝并事件的后果处理．然后利用MMC算法对存在理论分析解的5种特殊工况进行数值求解,模拟结果与理论解符合很好,证明MMC算法具有良好的计算精度和较低的计算代价．最后分析了不同类型的凝并核对于凝并过程的影响,常凝并核和连续区布朗凝并核对小颗粒影响大一些,而线性凝并核和二次方凝并核对大颗粒影响大一些．     

利用随机模拟方法求解第二类积分方程

给出一种求解第二类Fredholm和Volterra积分方程的数值算法,算法在数值积分技术的基础上使用Monte Carlo随机模拟方法求积分方程的近似解.通过数值例子证明了该算法是有效的.     

Monte Carlo EM加速算法

EM算法是近年来常用的求后验众数的估计的一种数据增广算法, 但由于求出其E步中积分的显示表达式有时很困难, 甚至不可能, 限制了其应用的广泛性. 而Monte Carlo EM算法很好地解决了这个问题, 将EM算法中E步的积分用Monte Carlo模拟来有效实现, 使其适用性大大增强. 但无论是EM算法, 还是Monte Carlo EM算法, 其收敛速度都是线性的, 被缺损信息的倒数所控制, 当缺损数据的比例很高时, 收敛速度就非常缓慢. 而Newton-Raphson算法在后验众数的附近具有二次收敛速率. 本文提出Monte Carlo EM加速算法, 将Monte Carlo EM算法与Newton-Raphson算法结合, 既使得EM算法中的E步用Monte Carlo模拟得以实现, 又证明了该算法在后验众数附近具有二次收敛速度. 从而使其保留了Monte Carlo EM算法的优点, 并改进了Monte Carlo EM算法的收敛速度. 本文通过数值例子, 将Monte Carlo EM加速算法的结果与EM算法、Monte Carlo EM算法的结果进行比较, 进一步说明了Monte Carlo EM加速算法的优良性.     

基于信用风险修正的多阶段银行资产组合优化模型

从资产组合管理角度出发,用信用风险修正的方法对企业信用等级阈值进行修正,同时考虑商业银行持续经营的特点,将修正后的信用风险引入到多阶段的模型当中去,建立一个基于信用风险修正的多阶段银行资产组合优化模型.针对该模型的特点,给出了把Monte Carlo模拟的动态算法和改进粒子群的多阶段算法相结合求解方法.数值试验表明所建...     

线性互补问题的并行多分裂松弛迭代算法

运用矩阵多重分裂理论,同时考虑并行计算与松弛迭代法,得到一类求解线性互补问题的高效数值算法．当问题的系数矩阵为对角元为正的H-矩阵或对称半正定矩阵时,证明了算法的全局收敛性；该算法与已有算法相比,具有计算量小、计算速度快等特点,因而特别适于求解大规模问题．数值试验的结果说明了算法的有效性．     

基于有效Monte Carlo模拟的外汇期权组合非线性VaR模型

为了克服极小概率事件发生概率估计的困难,提出了把重要抽样技术发展到外汇期权组合非线性VaR模型中,估计出组合损失概率。为了进一步达到减少模拟估计误差目的,在重要抽样技术基础上使用分层抽样技术,进行更有效的Monte Carlo模拟。数值结果表明,重要抽样技术算法比常用Monte Carlo模拟法的计算效率更有效;而重要抽样技术和分层抽样技术相结合算法比重要抽样技术算法更有效地减少模拟所要估计的组合损失概率的方差,有着更高的计算效率。     

蒙特卡洛方法(Ⅰ)

蒙特卡洛(Monte Carlo)方法,是一类通过随机变量的统计试验、随机模拟,求解数学物理、工程技术问题近似解的数值方法.在文献中[1,5,29,34,36,39],也把这类方法叫做统计试验方法,随机模拟方法.     

电缆方程的多重网格并行代数法（英文）

致力于研究求解线性代数方程组的多重网格并行算法,该算法是基于构建矩阵序列的经典Runge-Stuben(RS)方法及其改进的并行修正独立集合(PMIS)方法的.展示了求解离散电缆方程式所得到的线性代数方程组的结果,而电缆方程是用作描述电信号传播的.在求解中用到了GPUPU技术.展示了模型问题在不同尺度的模拟区域上的数值结果.     

流体饱和多孔隙介质波动方程小波有限差分法

研究流体饱和多孔隙介质中波动方程的数值模拟．针对求解二维弹性波方程问题,提出小波有限差分法．该方法综合了小波多分辨分析计算灵活、计算效率高特性和有限差分易于实现的优点．数值模拟的结果显示,此方法对于求解流体饱和多孔隙介质方程的数值模拟是有效稳定的．     

MC方差减小技术在算术平均亚式外汇期权定价中的应用

建立了利率和汇率波动率均为随机情形下算术平均亚式外汇期权的定价模型.由于其定价问题求解十分困难,运用蒙特卡罗(Monte Carlo)方法并结合控制变量方差减小技术进行模拟,有效地减小了模拟方差,得到了期权定价问题的数值结果.     

Study and analysis of a two‐dimensional nonconservative fractional order aerosol transport equation

The shifted Legendre collocation method is used to solve the two‐dimensional fractional order aerosol equation with initial and boundary conditions. The solution profile of the equation is presented graphically for different cases. The important feature of the article is graphical exhibitions of the effect of the size of the aerosol particles and also the temporal derivative on the solution profile. The salient feature of the article is the demonstration of lower variation of mass concentration with the change in time level in fractional order systems than that in integer order system. A drive has been taken towards the tabular and pictorial presentations of a comparison of the numerical solution of our proposed method with an analytical solution of an existing problem through error analysis which conforms super‐linearly convergence rate of the proposed method to validate its efficiency and effectiveness.     

Convergence analysis of finite volume scheme for nonlinear aggregation population balance equation

In this work, we introduce the convergence analysis of the recently developed finite volume scheme to solve a pure aggregation population balance equation that is of substantial interest in many areas such as chemical engineering, aerosol physics, astrophysics, polymer science, pharmaceutical sciences, and mathematical biology. The notion of the finite volume scheme is to conserve total mass of the particles in the system by introducing weight in the formulation. The consistency of the finite volume scheme is also analyzed thoroughly as it is an influential factor. The convergence study of the numerical scheme shows second order convergence on uniform, nonuniform smooth (geometric) as well as on locally uniform meshes independent of the aggregation kernel. Moreover, the first‐order convergence is shown when the finite volume scheme is implemented on oscillatory and random meshes. In order to check the accuracy, the numerical experimental order of convergence is also computed for the physically relevant as well as analytically tractable kernels and validated against its analytical results.     

Can Data Recognize Its Parent Distribution?

This study is concerned with model selection of lifetime and survival distributions arising in engineering reliability or in the medical sciences. We compare various distributions—including the gamma, Weibull, and lognormal—with a new distribution called geometric extreme exponential. Except for the lognormal distribution, the other three distributions all have the exponential distribution as special cases. A Monte Carlo simulation was performed to determine sample sizes for which survival distributions can distinguish data generated by their own families. Two methods for decision are by maximum likelihood and by Kolmogorov distance. Neither method is uniformly best. The probability of correct selection with more than one alternative shows some surprising results when the choices are close to the exponential distribution.     

On gradient methods for maximum entropy regularizing retrieval of atmospheric aerosol particle size distribution function

The determination of the aerosol particle size distribution function using the particle spectrum extinction equation is an ill-posed integral equation of the first kind, since as is known, we are often faced with limited or insufficient observations in remote sensing and the observations are contaminated. Physically, the particle size distribution is always nonnegative, and we are often faced with incomplete data. Therefore, the concept of maximum entropy from information theory and statistic mechanics can be used to counteract this problem of missing or erroneous data. Therefore, in this paper, we study the maximum entropy based regularization model and gradient methods for solving the corresponding optimization problem. Numerical tests are made for synthetic aerosol data to show the efficiency and feasibility of the proposed algorithms. (© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim)     

用格子Boltzmann方法模拟非线性热传导方程

对具有非线性源项和非线性扩散项的热传导方程建立格子Boltzmann求解模型.在演化方程中增加了两个关于源项分布函数的微分算子,对演化方程实施Chapman-Enskog展开.通过对演化方程的进一步改进,恢复出具有高阶截断误差的宏观方程.对不同参数选取下的非线性热传导方程进行了数值模拟,数值解与精确解吻合得很好.该模型也可以用于同类型的其他偏微分方程的数值计算中.     

Numerical modeling of interaction between surface radiation and natural convection of atmospheric aerosol in presence of transverse magnetic field

After the era of industrialization, technology is developing daily since the last century. Urbanization, communication, and transportation have grown rapidly and simultaneously deforestation and volcanic eruptions take place on a large scale. As result every moment tons of foreign particles like soot, dust, ash, and bio-fuel contaminants are released into the atmosphere. These contaminants mix with air and various green house gases, form a blanket structure in atmosphere. This mixture of ultrafine particle suspension with atmospheric air is known as aerosol. In the present study, numerical simulations of hydrodynamic single cell buoyant convection of atmospheric aerosol sample enclosed within a gray enclosure in the presence of a transverse magnetic field and surface radiation is addressed. Flow of the aerosol over deserts and industrial belts is a practical example of such a condition, where the thermal radiation emanating from the surface, affects the flow mechanism of the aerosol transport. The emphasis of the present study is only on carbon-black solid particles of a size in the nanometer range present in atmospheric air. The aerosol is treated as nanofluid for the numerical simulation. A comprehensive study on the controlling parameters that affect the flow and heat transfer characteristics are delineated. The governing equations are solved using modified MAC method and SIMPLER algorithm has been used to solve pressure velocity coupling employing relaxation technique. The transport equation for surface radiation is solved using the net radiation method. The cross string method is used to evaluate the view factor. The most striking result is that the heat transfer rate increases with increase in the volume fraction of the carbon-black particles, which has an adverse effect on both the climate and living creatures. The results are presented in tabular and graphical form. The heat transfer and flow characteristics are depicted in the form of isotherms and streamlines revealing the physics of this complex phenomenon.     

Numerical solution of a heated subsidence mound problem in a porous medium

Two different numerical models are constructed to solve a two dimensional subsidence mound problem heated along the moving wet/dry interface. One numerical model is based on cartesian coordinates while the other is based on polar coordinates. In both approaches coordinate transformations are used that render the interface stationary. The problem involves a system of three coupled equations; an elliptic equation for a stream function, a parabolic equation for the temperature and a non-linear equation for the boundary location. Good agreement is found between the results of both methods. Graphic results are presented for the decay of a subsidence mound for different values of the various parameters in the model problem.     

梯度下降法在沉积物粒度分布拟合中的应用

结合实测数据,以三个对数正态分布函数的和函数为拟合函数,以梯度下降法为主要方法,对沉积物粒度分布进行了数据拟合,通过数值实验我们发现:利用梯度下降法可以有效地优化分布函数的各参数,实现拟合残差的稳步持续减小,具有良好的可操作性,拟合效果是令人满意的,它为我们进行数据拟合提供了一条新的思路,同时此方法也可以推广到解决其他极值问题.     

Application of gradient descent method to the sedimentary grain-size distribution fitting

Existence of a least squares solution for a sum of several weighted normal functions is proved. The gradient descent (GD) method is used to fit the measured data (i.e. the laser grain-size distribution of the sediments) with a sum of three weighted lognormal functions. The numerical results indicate that the GD method is not only easy to operate but also could effectively optimize the parameters of the fitting function with the error decreasing steadily. Meanwhile the overall fitting results are satisfactory. As a new way of data fitting, the GD method could also be used to solve other optimization problems.     

Similarity solution of the telford coagulation equation

A direct operational method is used to transform the collision integral in the Telford coagulation equation for aerosols to a polynomial in derivatives with respect to volume. This polynomial is approximated by the Padé approximation technique. This procedure reduces the integro-differential form to a partial differential equation which is solved by the similarity method. Exact as well as approximate expressions for the particle size distribution are obtained.