The decay rate for a fractional differential equation

We consider the fractional differential equation

     

Exponential sums on , II

We prove a vanishing theorem for the -adic cohomology of exponential sums on . In particular, we obtain new classes of exponential sums on that have a single nonvanishing -adic cohomology group. The dimension of this cohomology group equals a sum of Milnor numbers.

     

Infinite order differential operators in spaces of entire functions

Differential operators ?(Δ_θ,ω), where ? is an exponential type entire function of a single complex variable and Δ_θ,ω=(θ+ωz)D+zD², D=∂/∂z, , θ?0, , acting in the spaces of exponential type entire function are studied. It is shown that, for ω?0, such operators preserve the set of Laguerre entire functions provided the function ? also belongs to this set. The latter consists of the polynomials possessing real nonpositive zeros only and of their uniform limits on compact subsets of the complex plane . The operator exp(aΔ_θ,ω), a?0 is studied in more details. In particular, it is shown that it preserves the set of Laguerre entire functions for all . An integral representation of exp(aΔ_θ,ω), a>0 is obtained. These results are used to obtain the solutions to certain Cauchy problems employing Δ_θ,ω.     

发展在XLPE电缆绝缘内外侧的电树枝

以XLPE高压电力电缆内外侧绝缘中的电树枝特性为研究对象，通过分析电树枝引发与生长的统计实验规律和采用扫描电子显微镜分析发现，由于不同结晶状态的影响，电缆绝缘内外侧的电树枝特性存在很大的差异.引发于绝缘内侧电树枝引发时间短、生长速度快、电树枝形状具有多样性；起始于绝缘外侧的电树枝不仅引发时间长、生长速度极慢，而且电树枝形状(结构)比较单一.并对这两个位置电树枝的引发和生长机理进行了探讨. 关键词： 电树枝 结晶状态 统计规律 内侧和外侧绝缘层     

Growth kinetics of borided layers: Artificial neural network and least square approaches

The present study evaluates the growth kinetics of the boride layer Fe₂B in AISI 1045 steel, by means of neural networks and the least square techniques. The Fe₂B phase was formed at the material surface using the paste boriding process. The surface boron potential was modified considering different boron paste thicknesses, with exposure times of 2, 4 and 6 h, and treatment temperatures of 1193, 1223 and 1273 K. The neural network and the least square models were set by the layer thickness of Fe₂B phase, and assuming that the growth of the boride layer follows a parabolic law. The reliability of the techniques used is compared with a set of experiments at a temperature of 1223 K with 5 h of treatment time and boron potentials of 2, 3, 4 and 5 mm. The results of the Fe₂B layer thicknesses show a mean error of 5.31% for the neural network and 3.42% for the least square method.     

Multiple nano-TiO2 layers to prevent dye/nano-TiO2 from photodegradation under a UV-exposure environment

A device of multiple nano-TiO₂ layers was proposed and fabricated to prevent a dye/nano-TiO₂ region from serious photo-degradation. In this device, the top of the dye/TiO₂ region was designed to be coated using sol-gel nano-TiO₂ thin films to shield UV irradiation from the photo-degradation effect. The sol-gel TiO₂ was prepared in a low temperature (75 °C) and verified as nano-sized particles and an anatase crystalline structure. Different devices of the multi-layer samples fabricated using different compositions of nano-TiO₂ were produced and exposed for UV irradiation tests. Results show that the presence of the sol-gel TiO₂ films coated on top of the dye/TiO₂ region can significantly alleviate the dye photo-degradation under UV irradiation. This multi-layer device can effectively improve the photo-stability of the dye/TiO₂ region in a UV-exposure environment.     

Dynamics of Cohen-Grossberg neural networks with variable and distributed delays

This paper is concerned with the analysis problem for the exponential stability of a class of Cohen-Grossberg neural networks with variable and distributed delays. Some sufficient conditions ensuring the existence, uniqueness and exponential stability of the equilibrium point are obtained by employing Brouwer’s fixed-point theorem and by applying the inequality technique. In the results, we do not assume that the activation function satisfies the boundedness and the Lipschitz condition. Three numerical examples are given to show the effectiveness of the obtained results.     

An implementation of the exponential time differencing scheme to the magnetohydrodynamic equations in a spherical shell

Over the last decade there has been renewed interest in applying exponential time differencing (ETD) time stepping schemes to the solution of stiff systems. In this paper, we present an implementation of such a scheme to the fully spectral solution of the incompressible magnetohydrodynamic equations in a spherical shell. One problem associated with ETD schemes is the accurate calculation of the necessary matrices; we implement and discuss in detail a variety of different methods including direct computation, contour integration, spectral expansions and recurrence relations. We compare the accuracy of six different second-order methods in determining the evolution of a three-dimensional magnetic field under the action of a prescribed time-dependent flow of electrically conducting fluid, and find that for the timestep restriction imposed by the nonlinear terms, ETD methods are no more accurate than linearly implicit methods which have the significant advantage of being easier to implement. However, ETD methods are more readily extendable than those which are linearly implicit and will become much more advantageous at higher order.     

Asymptotic exponential cones of Metzler matrices and their use in the solution of an algebraic problem

The aim of this paper is that of investigating the asymptotic exponential cone of a single Metzler matrix, introduced in [23], and of defining and analysing the new concept of asymptotic exponential cone of a family of Metzler matrices (along a certain direction). These results will provide necessary and/or sufficient conditions for the solvability of an interesting algebraic problem that arises in the context of continuous-time positive switched systems and, specifically, in the investigation of the reachability property [21,22,25].     

Convergence of the Magnus Series

The Magnus series is an infinite series which arises in the study of linear ordinary differential equations. If the series converges, then the matrix exponential of the sum equals the fundamental solution of the differential equation. The question considered in this paper is: When does the series converge? The main result establishes a sufficient condition for convergence, which improves on several earlier results.