Vol 21, No 6 (2015)

Existence and behavior of oscillatory solutions to nonlinear equations with regular and singular power nonlinearity are investigated. In particular, the existence of oscillatory solutions is proved for the equation y(n) + P(x; y; y ′ ; : : : ; y(n−1))|y|k sign y = 0; n > 2; k ∈ R; k > 1; P ̸= 0; P ∈ C(Rn+1): A criterion is formulated for oscillation of all solutions to the quasilinear even-order differential equation y(n) + nΣ−1 i=0 aj(x) y(i) + p(x) |y|ksigny = 0; p ∈ C(R); aj ∈ C(R); j = 0; : : : ; n − 1; k > 1; n = 2m; m ∈ N; which generalizes the well-known Atkinson’s and Kiguradze’s criteria. The existence of quasi-periodic solutions is proved both for regular (k > 1) and singular (0 < k < 1) nonlinear equations y(n) + p0 |y|ksigny = 0; n > 2; k ∈ R; k > 0; k ̸= 1; p0 ∈ R; with (−1)np0 > 0: A result on the existence of periodic oscillatory solutions is formulated for this equation with n = 4; k > 0; k ̸= 1; p0 < 0:

The solution of Riquier’s problem - the problem of finding in n-dimensional ball of solving k + 1 - harmonic equation for given values on the boundary of the desired solution u and powers of the Laplacian from one to k inclusive of this decision is obtained. The first part provides an exact statement of the problem, the main result (form of the solution of it), and the idea of this proof is stated. The second part introduces a family of some differential and integral operators in the space of harmonic functions in the ball used in the proof of the main result; some properties of these operators are set. The content of the third part is the proof of the main result. It is based on the properties of operators introduced in the second part.

Consider the second-order differential equation of Emden - Fowler type with negative potential y′′ - p (x, y, y′) |y|sgn y = 0: The function p (x; y; y′) is assumed positive, continuous, and Lipschitz continuous in y, y′: In the case ofsingular nonlinearity (0 < k < 1) the solutions to above equation can behavein a special way not only near the boundaries of their domains but also near internal points of the domains. This is why a notion of maximally uniquely extended solutions is introduced. Asymptotic classification of non-extensible solutions to above equation in case of regular nonlinearity (k > 1) and classification of maximally uniquely extended solutions to above equation in case of singular nonlinearity (0 < k < 1) are obtained.

The inverse problem of finding initial conditions and the right-hand side had been studied for the inhomogeneous heat equation on the basis of formulas for the solution of the first initial-boundary value problem. A criterion of uniqueness of solution of the inverse problem for finding the initial condition was found with Spectral analysis. The right side of the heat equation is represented as a product of two functions, one of which depends on the spatial coordinates and the other from time. In one task, along with an unknown solution is sought factor on the right side, depending on the time, and in another - a factor that depends on the spatial coordinates. For these tasks, we prove uniqueness theorems, the existence and stability of solution.

At the proof of a classical Marсhaud inequality for equidistant moduli of continuity of the highest degree the reduction of their definition for arbitrary sign of a step of a finite difference to positive values of this step is used. In case of moduli of continuity with a weight such reduction reduces definitions of moduli of continuity to restriction. Consequently for determination of properties of moduli of continuity with a weight other approach of reasoning is required. Unlike usual weight signsensitive weight allows to consider not only an absolute value of an increment of function, but also a sign of this increment. In the work for metrics with signsensitive weight an analogue of Marchaud inequality on estimation of modulus of continuity of given degree over modulus of continuity of a higher degree is obtained.

The work is devoted to the numerical investigation of the generalized Hoff model. Hoff equation models the dynamics of buckling construction of I-beams under a constant load. Result of existence and uniqueness of solution to the Showalter - Sidorov problem for the investigated model is formulated. This equation is a semilinear Sobolev type equation. Sobolev type equations constitute a vast area of non-classical equations of mathematical physics. Based on the theoretical results there was developed the algorithm of numerical solution of the problem.

Uncertainty principles of harmonic analysis and their analogues for finite abelian groups are considered in the paper. Special attention is paid to the recent results of T. Tao and coauthors about cyclic groups of prime order. It is shown, that indicator functions of subgroups of finite Abelian groups are analogues of Gaussian functions. Finite-dimensional version of Poisson summation formula is proved. Opportunities of application of these results for reconstruction of discrete signals with incomplete number of coefficients are suggested. The principle of partial isometric whereby we can determine the minimum number of measurements for stable recovery of the signal are formulated.

A third order Emden - Fowler type equation is considered. Existence of solution with given number of zeros on given interval is proved. This theorem extends previous results, related to Emden - Fowler type equation with constant coefficient, in case of variable coefficient.

The paper deals with the solutions of ordinary differential semi-linear equa- tion, the coefficients of which depend on several real parameters. If the coefficient is chosen so that the equation does not contain the first-order derivative of the unknown function, it will be the case of Emden - Fowler equation. Asymp- totic behavior of Emden - Fowler equation solutions at infinity is described in the book of Richard Bellman. The equations with the first-order derivative, considered in this work, erase in some problems for elliptic partial differential equations in unbounded domains. The sign of the coefficient in first-order deriva- tive term essentially influences on the description of solutions. Partly the result of this paper can be obtained from the works of I.T. Kiguradze. In present work we use lemmas about the behavior of solutions of the linear equations with a strongly (weakly) increasing potential. The paper deals with the solutions of ordinary differential semi-linear equa- tion, the coefficients of which depend on several real parameters. If the coefficient is chosen so that the equation does not contain the first-order derivative of the unknown function, it will be the case of Emden - Fowler equation. Asymp- totic behavior of Emden - Fowler equation solutions at infinity is described in the book of Richard Bellman. The equations with the first-order derivative, considered in this work, erase in some problems for elliptic partial differential equations in unbounded domains. The sign of the coefficient in first-order deriva- tive term essentially influences on the description of solutions. Partly the result of this paper can be obtained from the works of I.T. Kiguradze. In present work we use lemmas about the behavior of solutions of the linear equations with a strongly (weakly) increasing potential.

The algorithm of numerical integration of a task of Cauchy for the equations of the movement of self-oscillatory systems of Thomson type is offered. The algorithm is based on the use of samples of impulse response of linear resonant system as discretization sequences at the transition to the discrete time in the integral form of the equations of motion. Estimates of an error of numerical decisions are given. Transformation of finite difference algorithm in object of nonlinear dynamics in discrete time is discussed. Version of discrete mapping of Van der Pol oscillator is proposed. The algorithm of numerical integration of a task of Cauchy for the equations of the movement of self-oscillatory systems of Thomson type is offered. The algorithm is based on the use of samples of impulse response of linear resonant system as discretization sequences at the transition to the discrete time in the integral form of the equations of motion. Estimates of an error of numerical decisions are given. Transformation of finite difference algorithm in object of nonlinear dynamics in discrete time is discussed. Version of discrete mapping of Van der Pol oscillator is proposed. The algorithm of numerical integration of a task of Cauchy for the equations of the movement of self-oscillatory systems of Thomson type is offered. The algorithm is based on the use of samples of impulse response of linear resonant system as discretization sequences at the transition to the discrete time in the integral form of the equations of motion. Estimates of an error of numerical decisions are given. Transformation of finite difference algorithm in object of nonlinear dynamics in discrete time is discussed. Version of discrete mapping of Van der Pol oscillator is proposed.

The article deals with two-dimensional and three-dimensional Ising models with the long-range spin interactions. The intensity of interaction between the spins relies decreasing with distance r as a power law r-d-σ with dimensional d and parameter σ. The research are conducted by Monte-Carlo method with Metropolis algorithm using parallel computing techniques. On the basis of nu- merical simulation the dependence of the phase transition temperature on the parameter σ is found. It is shown that at phase transition temperature decreases with increasing σ.

In the article methodical and technical solutions which essentially expand capabilities of creation of the electronic intelligent educational resources contain- ing mathematical notation of any level of complexity and graphics illustrations of all types and categories are discussed. Base units of program modules, key constructions of codes, functions and options of language of the system of com- puter algebra Mathematica are explained. Main rules of preparation of freely distributed interactive program applications of CDF format are noted. Exam- ples from practice of preparation of teaching materials of discipline ”Computer Graphics” are given. User interface and results of execution of program modules are illustrated.