Model description of surface charging during ultra-fast pulsed laser ablation of materials

We present a model describing the dynamical mechanisms responsible for generating fast ion ejection under ultra-short pulsed laser irradiation. The model is based on a simplified drift–diffusion approach describing the evolution of the laser-generated charge carriers, their transport, and the electric field generated as a result of quasi-neutrality breaking in the irradiated target. The importance of different processes in generating the non-thermal material-ejection mechanisms is discussed. A common frame is applied to dielectrics, semiconductors, and metals and different dynamical behaviour is observed. The modelling results are in good agreement with fs pump–probe studies and measurements of the velocity distributions of the emitted ions.     

Non-linear relativistic diffusions

We obtain a non-linear generalization of the relativistic diffusion. We discuss diffusion equations whose non-linearity is a consequence of quantum statistics. We show that the assumptions of the relativistic invariance and an interpretation of the solution as a probability distribution substantially restrict the class of admissible non-linear diffusion equations. We consider relativistic invariant as well as covariant frame-dependent diffusion equations with a drift. In the latter case we show that there can exist stationary solutions of the diffusion equation besides the equilibrium solution corresponding to the quantum or Tsallis distributions. We define the relative entropy as a function of the diffusion probability and prove that it is monotonically decreasing in time when the diffusion tends to equilibrium. We discuss its relation to the thermodynamic behavior of diffusing particles.     

Theory of hot-electron quantum diffusion in semiconductors

In the linear response regime close to equilibrium, the fluctuation-dissipation theorem relates linear transport coefficients via the well-known Green–Kubo or Einstein relation. The latter embodies a deep connection between fluctuations causing diffusion and dissipation, which are responsible for a finite mobility. Far from equilibrium, however, the Einstein relation is no longer valid so that both the mobility and diffusivity gain their own physical integrity. Consequently, beyond a linear response, both quantities have to be described by different approaches. Unfortunately, there is a strong imbalance of research activities devoted to the study of both transport mechanisms in semiconductors. On one hand, the rich physics of high-field quantum drift in semiconducting structures has a long history and has reached a high level of sophistication. On the other hand, there are only comparatively few and unsystematic studies that cover quantum diffusion of carriers under high-field conditions. This review aims at reducing this gap by presenting a unified approach to quantum drift and quantum diffusion. Starting from a semi-phenomenological basis, a quantum theory of transport coefficients is developed for one- as well as multi-band models. Physical implications are illustrated by selected applications whereby the quantum character of the approach is emphasized. Furthermore, the basic unified treatment of transport coefficients is extended by accounting for the two-time dependence of one-particle correlation functions in quantum statistics. As an application, a phononless transport mechanism is identified, which solely originates from the double-time nature of the evolution. Finally, additional examples are presented that illustrate the important role played by quantum diffusion in semiconductor physics.     

Ambipolar diffusion and drift in computational weakly-ionized plasmadynamics

Modeling of ambipolar diffusion and drift taking place within a weakly-ionized fluid can lead to some convergence difficulties when the ion conservation equation and the electric field potential equation are solved consecutively. A novel formulation of the ion flow rate is proposed here that reduces the computing effort to reach convergence by a factor of 10 or more. It is shown that by recasting the ion flow rate in terms of drift and ambipolar diffusion components, the sensitivity to the electric field is reduced hence alleviating the stiffness of the system of equations and permitting significantly faster convergence. What makes the method particularly appealing is that (i) it yields faster convergence without affecting the accuracy of the converged solution and (ii) it is not restricted to specific discretization or relaxation schemes and can hence be readily implemented in existing flow solvers. Because it is developed in general form (i.e. applicable to a multicomponent plasma in the simultaneous presence of electric current and magnetic and electric fields), the method is notably well-suited to simulate ambipolar diffusion within ionized multi-species flow solvers and is recommended for all flowfields as long as the plasma remains weakly-ionized and quasi-neutral.     

Optimal Non-reversible Linear Drift for the Convergence to Equilibrium of a Diffusion

We consider non-reversible perturbations of reversible diffusions that do not alter the invariant distribution and we ask whether there exists an optimal perturbation such that the rate of convergence to equilibrium is maximized. We solve this problem for the case of linear drift by proving the existence of such optimal perturbations and by providing an easily implementable algorithm for constructing them. We discuss in particular the role of the prefactor in the exponential convergence estimate. Our rigorous results are illustrated by numerical experiments.     

Estimating parameters in stochastic systems: A variational Bayesian approach

This work is concerned with approximate inference in dynamical systems, from a variational Bayesian perspective. When modelling real world dynamical systems, stochastic differential equations appear as a natural choice, mainly because of their ability to model the noise of the system by adding a variation of some stochastic process to the deterministic dynamics. Hence, inference in such processes has drawn much attention. Here a new extended framework is derived that is based on a local polynomial approximation of a recently proposed variational Bayesian algorithm. The paper begins by showing that the new extension of this variational algorithm can be used for state estimation (smoothing) and converges to the original algorithm. However, the main focus is on estimating the (hyper-) parameters of these systems (i.e. drift parameters and diffusion coefficients). The new approach is validated on a range of different systems which vary in dimensionality and non-linearity. These are the Ornstein-Uhlenbeck process, the exact likelihood of which can be computed analytically, the univariate and highly non-linear, stochastic double well and the multivariate chaotic stochastic Lorenz ’63 (3D model). As a special case the algorithm is also applied to the 40 dimensional stochastic Lorenz ’96 system. In our investigation we compare this new approach with a variety of other well known methods, such as the hybrid Monte Carlo, dual unscented Kalman filter, full weak-constraint 4D-Var algorithm and analyse empirically their asymptotic behaviour as a function of observation density or length of time window increases. In particular we show that we are able to estimate parameters in both the drift (deterministic) and the diffusion (stochastic) part of the model evolution equations using our new methods.     

Anomalous diffusion for a correlated process with long jumps

We discuss diffusion properties of a dynamical system, which is characterised by long-tail distributions and finite correlations. The particle velocity has the stable Lévy distribution; it is assumed as a jumping process (the kangaroo process) with a variable jumping rate. Both the exponential and the algebraic form of the covariance–defined for the truncated distribution–are considered. It is demonstrated by numerical calculations that the stationary solution of the master equation for the case of power-law correlations decays with time, but a simple modification of the process makes the tails stable. The main result of the paper is a finding that–in contrast to the velocity fluctuations–the position variance may be finite. It rises with time faster than linearly: the diffusion is anomalously enhanced. On the other hand, a process which follows from a superposition of the Ornstein–Uhlenbeck–Lévy processes always leads to position distributions with a divergent variance which means accelerated diffusion.     

Generalized pricing formulas for stochastic volatility jump diffusion models applied to the exponential Vasicek model

Path integral techniques for the pricing of financial options are mostly based on models that can be recast in terms of a Fokker-Planck differential equation and that, consequently, neglect jumps and only describe drift and diffusion. We present a method to adapt formulas for both the path-integral propagators and the option prices themselves, so that jump processes are taken into account in conjunction with the usual drift and diffusion terms. In particular, we focus on stochastic volatility models, such as the exponential Vasicek model, and extend the pricing formulas and propagator of this model to incorporate jump diffusion with a given jump size distribution. This model is of importance to include non-Gaussian fluctuations beyond the Black-Scholes model, and moreover yields a lognormal distribution of the volatilities, in agreement with results from superstatistical analysis. The results obtained in the present formalism are checked with Monte Carlo simulations.     

Diffusion processes with singular drift fields

A class of stochastic differential equations with highly singular drift fields is considered. Using a purely probabilistic approach, we can show the unattainability of the nodal set. Moreover, a global existence and uniqueness theorem for diffusion processes with singular drift fields is established. The finite action condition of Carlen and Zheng can be modified. We relate our results to the diffusions which describe the time evolution of quantum systems in stochastic mechanics.     

Behavior of general one-dimensional diffusion processes

We develop simple rigorous techniques to estimate the behavior of general one-dimensional diffusion processes. Any one-dimensional diffusion process (with drift) can be mapped onto a symmetric diffusion through an explicit change of variable. For such processes we can estimate explicitly the diffusion exponent, the recurrence properties, and the large fluctuations. In a second part, we apply these results to different models (including the Sinaï random walk: diffusion in a random drift) and we show how the main features of the diffusion can be readily handled.     

Brownian Approximation and Monte Carlo Simulation of the Non-Cutoff Kac Equation

The non-cutoff Boltzmann equation can be simulated using the DSMC method, by a truncation of the collision term. However, even for computing stationary solutions this may be very time consuming, in particular in situations far from equilibrium. By adding an appropriate diffusion, to the DSMC-method, the rate of convergence when the truncation is removed, may be greatly improved. We illustrate the technique on a toy model, the Kac equation, as well as on the full Boltzmann equation in a special case.     

Kinetic Approach to Long time Behavior of Linearized Fast Diffusion Equations

We show that the rate of convergence towards the self-similar solution of certain linearized versions of the fast diffusion equation can be related to the number of moments of the initial datum that are equal to the moments of the self-similar solution at a fixed time. As a consequence, we find an improved rate of convergence to self-similarity in terms of a Fourier based distance between two solutions. The results are based on the asymptotic equivalence of a collisional kinetic model of Boltzmann type with a linear Fokker-Planck equation with nonconstant coefficients, and make use of methods first applied to the reckoning of the rate of convergence towards equilibrium for the spatially homogeneous Boltzmann equation for Maxwell molecules.     

论动态信息理论

本文综述了作者的研究成果.近十年,作者将现有静态统计信息理论拓展至动态过程,建立了以表述动态信息演化规律的动态信息演化方程为核心的动态统计信息理论.基于服从随机性规律的动力学系统(如随机动力学系统和非平衡态统计物理系统)与遵守确定性规律的动力学系统(如电动力学系统)的态变量概率密度演化方程都可看成是其信息符号演化方程,推导出了动态信息(熵)演化方程.它们表明:对于服从随机性规律的动力学系统,动态信息密度随时间的变化率是由其在系统内部的态变量空间和传递过程的坐标空间的漂移、扩散和耗损三者引起的,而动态信息熵密度随时间的变化率则是由其在系统内部的态变量空间和传递过程的坐标空间的漂移、扩散和产生三者引起的.对于遵守确定性规律的动力学系统,动态信息(熵)演化方程与前者的相比,除动态信息(熵)密度在系统内部的态变量空间仅有漂移外,其余皆相同.信息和熵已与系统的状态和变化规律结合在一起,信息扩散和信息耗损同时存在.当空间噪声可略去时,将会出现信息波.若仅研究系统内部的信息变化,动态信息演化方程就约化为与表述上述动力学系统变化规律的动力学方程相对应的信息方程,它既可看成是表述动力学系统动态信息的演化规律,亦可看成是动力学系统的变化规律都可由信息方程表述.进而给出了漂移和扩散信息流公式、信息耗散率公式和信息熵产生率公式及动力学系统退化和进化的统一信息表述公式.得到了反映信息在传递过程中耗散特性的动态互信息公式和动态信道容量公式,它们在信道长度和信号传递速度之比趋于零的极限情况下变为现有的静态互信息公式和静态信道容量公式.所有这些新的理论公式和结果都是从动态信息演化方程统一推导出的.     

Convergence from Boltzmann to Landau Processes with Soft Potential and Particle Approximations

Our aim in this paper is to show how a probabilistic interpretation of the Boltzmann and Landau equations gives a microscopic understanding of these equations. We firstly associate stochastic jump processes with the Boltzmann equations we consider. Then we renormalize these equations following asymptotics which make prevail the grazing collisions, and prove the convergence of the associated Boltzmann jump processes to a diffusion process related to the Landau equation. The convergence is pathwise and also implies a convergence at the level of the partial differential equations. The best feature of this approach is the microscopic understanding of the transition between the Boltzmann and the Landau equations, by an accumulation of very small jumps. We deduce from this interpretation an approximation result for a solution of the Landau equation via colliding stochastic particle systems. This result leads to a Monte-Carlo algorithm for the simulation of solutions by a conservative particle method which enables to observe the transition from Boltzmann to Landau equations. Numerical results are given.     

多维梯度折射率介质内辐射传递的扩散近似

推导多维梯度折射率介质内稳态辐射传递的扩散近似方程.使用有限元法对扩散近似进行离散和求解,利用两个二维半透明介质的稳态辐射传递问题验证该扩散近似的精度及适用性.算例考虑介质为均匀折射率及梯度折射率两种情况.利用扩散近似分别求解辐射平衡时的边界热流、介质内温度场分布,并与辐射传递方程的求解结果进行对比分析.结果表明:介质折射率变化、散射特性、光学厚度及散射反照率均直接影响扩散近似的精度;在光学厚及强散射条件下,该扩散近似可以作为一种快速算法应用于梯度折射率介质稳态辐射传递的求解.     

Dust Self‐Organized Structures I. Role of Ion Drag and Ion Diffusion on Screened Grains

Self‐organized dust structures are investigated using a Gurevich‐Parker model for non‐linear dust screening. The non‐linear dust drag coefficients and non‐linear diffusion coefficients are calculated numerically as functions of nonlinear parameter for screening, dust density and ion flux drift velocity. Nonlinear ion dust drag inside the structures creates an electric field with potential well for ions at the structure center. The equilibrium dust structures confine both the dust grains and the plasma particles, have a finite size and have inside an enhanced dust and ion densities. The necessary conditions for existence of equilibrium dust structures are found. The equilibrium dust structures are determined by two global parameters related to the external plasma flux and to the power of ionization. The equilibrium exist only in a restricted phase space of these two parameters and depends on the the drag coefficient at the structure center. The equilibrium requirements are found using non‐linear drag coefficient calculated numerically. It is shown that this phase space area can be broad but it is systematically decreasing with an increase of the ionization rate. It is found that equilibrium exists for dust structures with large dust and ion density concentration at the center and that for these structures the ion diffusion is strongly suppressed by ion scattering on non‐linearly screened grains. The results of the theory can be used to interpret the recently observed compact dust structures in micro‐gravity experiments and can provide some recommendations for future micro‐gravity experiments in spherical chambers (© 2011 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)     

Adaptive Sampling of Large Deviations

We introduce and test an algorithm that adaptively estimates large deviation functions characterizing the fluctuations of additive functionals of Markov processes in the long-time limit. These functions play an important role for predicting the probability and pathways of rare events in stochastic processes, as well as for understanding the physics of nonequilibrium systems driven in steady states by external forces and reservoirs. The algorithm uses methods from risk-sensitive and feedback control to estimate from a single trajectory a new process, called the driven process, known to be efficient for importance sampling. Its advantages compared to other simulation techniques, such as splitting or cloning, are discussed and illustrated with simple equilibrium and nonequilibrium diffusion models.     

玉米品种近红外光谱鉴别技术中的参数漂移问题研究

以13个玉米品种鉴别为研究对象,提出了一种解决光谱仪参数漂移问题的有效方法。使用同一台光谱仪分不同时间重复采集数据,用一天数据建模,其余测试,发现不同时间采集的数据有较大偏移,严重时正确识别率仅为7.69%。为此,提出一种有监督学习特征提取的多天联合建模方法,首先挑选具有代表性的多个时间段样本数据共同组成建模集,其次采用PLS+LDA特征提取算法,提取出与仪器参数漂移无关的品种特征信息, 然后采用BPR方法建立品种鉴别模型。实验结果表明,该方法对于不同时间数据的偏移均能有较好的校正效果,得到较高的识别率和稳定性。     

The stress-driven migration of point defects to a slowly moving crack

The migration kinetics of point defects near a slowly moving brittle crack are studied under the condition of pure drift. In the pure-drift approximation it is assumed that the point-defect flow in the vicinity of a crack tip is dominated by the elastic interaction between the stress field of the crack and a point defect and that concentration gradient effects can be neglected. While such a pure-drift approach has been shown to be useful to calculate the short-time diffusion kinetics of impurity-induced subcritical crack growth, previous applications are based on the drift solutions for a stationary crack. In the present paper, the first-order drift diffusion equation for a slowly moving crack at uniform velocity is solved. This yields the flow lines of the point defects and the impurity segregation rate directly in terms of the crack growth rate. The flow line patterns reveal important insights with respect to the point-defect migration kinetics near a steadily advancing crack. Although the calculation is entirely elastic, it is shown that the present drift model maintains some relevance also in the presence of a plastic zone ahead of the crack tip.