Vol 56, No 3 (2016)
- Year: 2016
- Articles: 16
- URL: https://journals.rcsi.science/0965-5425/issue/view/11073
Article
On the principal and strictly particular solutions to infinite systems
Abstract
The concepts of the principal solution to infinite systems of linear algebraic equations and the reduction method are defined more precisely. The principal solution, if it exists, is a strictly particular solution to the infinite system. If the reduction method is convergent, then it necessarily converges to Kramer’s determinant; however, Kramer’s determinant is not always a solution to the infinite system. To confirm the obtained results, analytical and numerical solutions of specific infinite system are considered.
Analogue of Newton–Cotes formulas for numerical integration of functions with a boundary-layer component
Abstract
The numerical integration of functions with a boundary-layer component whose derivatives are not uniformly bounded is investigated. The Newton–Cotes formulas as applied to such functions can lead to significant errors. An analogue of Newton–Cotes formulas that is exact for the boundary-layer component is constructed. For the resulting formula, an error estimate that is uniform with respect to the boundary-layer component and its derivatives is obtained. Numerical results that agree with the error estimates are presented.
Reconstruction of random-disturbance amplitude in linear stochastic equations from measurements of some of the coordinates
Abstract
The problem of reconstructing the unknown amplitude of a random disturbance in a linear stochastic differential equation is studied in a fairly general formulation by applying dynamic inversion theory. The amplitude is reconstructed using discrete information on several realizations of some of the coordinates of the stochastic process. The problem is reduced to an inverse one for a system of ordinary differential equations satisfied by the elements of the covariance matrix of the original process. Constructive solvability conditions in the form of relations on the parameters of the system are discussed. A finite-step software implementable solving algorithm based on the method of auxiliary controlled models is tested using a numerical example. The accuracy of the algorithm is estimated with respect to the number of measured realizations.
Numerical algorithm for solving mathematical programming problems with a smooth surface as a constraint
Abstract
A numerical algorithm for minimizing a convex function on a smooth surface is proposed. The algorithm is based on reducing the original problem to a sequence of convex programming problems. Necessary extremum conditions are examined, and the convergence of the algorithm is analyzed.
Application of linear programming techniques for controlling linear dynamic plants in real time
Abstract
The problem of controlling a linear dynamic plant in real time given its nondeterministic model and imperfect measurements of the inputs and outputs is considered. The concepts of current distributions of the initial state and disturbance parameters are introduced. The method for the implementation of disclosable loop using the separation principle is described. The optimal control problem under uncertainty conditions is reduced to the problems of optimal observation, optimal identification, and optimal control of the deterministic system. To extend the domain where a solution to the optimal control problem under uncertainty exists, a two-stage optimal control method is proposed. Results are illustrated using a dynamic plant of the fourth order.
On a class of optimal control problems with distributed and lumped parameters
Abstract
The optimal control of moving sources governed by a parabolic equation and a system of ordinary differential equations with initial and boundary conditions is considered. For this problem, an existence and uniqueness theorem is proved, sufficient conditions for the Fréchet differentiability of the cost functional are established, an expression for its gradient is derived, and necessary optimality conditions in the form of pointwise and integral maximum principles are obtained.
Analysis of stability boundaries of satellite’s equilibrium attitude in a circular orbit
Abstract
An asymmetric satellite equipped with control momentum gyroscopes (CMGs) with the center of mass of the system moving uniformly in a circular orbit was considered. The stability of a relative equilibrium attitude of the satellite was analyzed using Lyapunov’s direct method. The Lyapunov function V is a positive definite integral of the total energy of the perturbed motion of the system. The asymptotic stability analysis of the stationary motion of the conservative system was based on the Barbashin–Krasovskii theorem on the nonexistence of integer trajectories of the set \(\dot V\), which was obtained using the differential equations of motion of the satellite with CMGs. By analyzing the sign definiteness of the quadratic part of V, it was found earlier by V.V. Sazonov that the stability region is described by four strict inequalities. The asymptotic stability at the stability boundary was analyzed by sequentially turning these inequalities into equalities with terms of orders higher than the second taken into account in V. The sign definiteness analysis of the inhomogeneous function V at the stability boundary involved a huge amount of computations related to the multiplication, expansion, substitution, and factorization of symbolic expressions. The computations were performed by applying a computer algebra system on a personal computer.
Justification of the Galerkin method for hypersingular equations
Abstract
The paper presents a theoretical study of hypersingular equations of the general form for problems of electromagnetic-wave diffraction on open surfaces of revolution. Justification of the Galerkin is given. The method is based on the separation of the principal term and its analytic inversion. The inverse of the principal operator is completely continuous. On the basis of this result, the equivalence of the initial equation to a Fredholm integral equation of the second kind is proven. An example of numerical solution with the use of Chebyshev polynomials of the second kind is considered.
Long-time convergence of numerical approximations for 2D GBBM equation
Abstract
We study the long-time behavior of the finite difference solution to the generalized BBM equation in two space dimensions with dirichlet boundary conditions. The unique solvability of numerical solution is shown. It is proved that there exists a global attractor of the discrete dynamical system. Finally, we obtain the long-time stability and convergence of the difference scheme. Our results show that the difference scheme can effectively simulate the infinite dimensional dynamical systems. Numerical experiment results show that the theory is accurate and the schemes are efficient and reliable.
Track method for the calculation of plasma heating by charged thermonuclear reaction products for axisymmetric flows
Abstract
Integral formulas for the three-dimensional case that give the plasma heating rate per unit volume are obtained using the track method and by integrating the well-known Cauchy problem for the steady-state homogeneous kinetic equation in the Fokker–Planck approximation in the absence of diffusion of the distribution function in the velocity space and under the condition that the velocity of the produced particles is independent on the direction of their escape. It is shown that both integral formulas are equivalent and, in the case of space homogeneous coefficients, turn into the model of local plasma heating away from the domain boundary. In addition to the known direct track method, the inverse method based on the approximation of the integral formula is developed. It is shown that the accuracy of the direct method is significantly decreased in the vicinity of the symmetry axis for not very fine angular grids. In the inverse method, the accuracy is not lost. It is shown that the computational cost of the inverse method can be significantly reduced without the considerable reduction of the computation accuracy.
Numerical solution of the equilibrium problem for a membrane with embedded rigid inclusions
Abstract
The equilibrium problem for a membrane containing a set of volume and thin rigid inclusions is considered. A solution algorithm reducing the original problem to a system of Dirichlet ones is proposed. Several examples are presented in which the problem is solved numerically by applying the finite element method.
Effective solving of three-dimensional gas dynamics problems with the Runge-Kutta discontinuous Galerkin method
Abstract
In this paper we present the Runge-Kutta discontinuous Galerkin method (RKDG method) for the numerical solution of the Euler equations of gas dynamics. The method is being tested on a series of Riemann problems in the one-dimensional case. For the implementation of the method in the three-dimensional case, a DiamondTorre algorithm is proposed. It belongs to the class of the locally recursive non-locally asynchronous algorithms (LRnLA). With the help of this algorithm a significant increase of speed of calculations is achieved. As an example of the three-dimensional computing, a problem of the interaction of a bubble with a shock wave is considered.
Unsteady rarefied gas flow in a microchannel driven by a pressure difference
Abstract
The kinetic S-model is used to study the unsteady rarefied gas flow through a plane channel between two parallel infinite plates. Initially, the gas is at rest and is separated by the plane x = 0 with different pressure values on opposite sides. The gas deceleration effect of the channel walls is studied depending on the degree of gas rarefaction and the initial pressure drop, assuming that the molecules are diffusely reflected from the boundary. The decay of the shock wave and the disappearance of the uniform flow region behind the shock wave are monitored. Special attention is given to the gas mass flux through the cross section at x = 0, which is computed as a function of time. The asymptotic behavior of the solution at unboundedly increasing time is analyzed. The kinetic equation is solved numerically by applying a conservative finite-difference method of second-order accuracy in space.
Inverse problem of determining parameters of inhomogeneity of a body from acoustic field measurements
Abstract
This work is devoted to development of methods for solving inverse problems in acoustics. Propagation of an acoustic field in a body located in the free space is considered. In the inverse problem, an iterative method for reconstructing the parameters of inhomogeneity of a body from a known acoustic field is applied. The theorem on convergence of the method is proven. Numerical results for inhomogeneous bodies of complex form are presented.
On the complexity of some quadratic Euclidean 2-clustering problems
Abstract
Some problems of partitioning a finite set of points of Euclidean space into two clusters are considered. In these problems, the following criteria are minimized: (1) the sum over both clusters of the sums of squared pairwise distances between the elements of the cluster and (2) the sum of the (multiplied by the cardinalities of the clusters) sums of squared distances from the elements of the cluster to its geometric center, where the geometric center (or centroid) of a cluster is defined as the mean value of the elements in that cluster. Additionally, another problem close to (2) is considered, where the desired center of one of the clusters is given as input, while the center of the other cluster is unknown (is the variable to be optimized) as in problem (2). Two variants of the problems are analyzed, in which the cardinalities of the clusters are (1) parts of the input or (2) optimization variables. It is proved that all the considered problems are strongly NP-hard and that, in general, there is no fully polynomial-time approximation scheme for them (unless P = NP).