Vol 224 (2023)
Статьи
Variation optimality condition of a boundary control in a composite model of linear differential equations of different types
Abstract
A linear optimal control problem of a system of differential equations with partial derivatives of the kinetic-diffusion type is considered. The controlled boundary condition is determined as a solution of a linear ordinary differential equation. Problems of this type arise when controlling the dynamics of populations, taking into account the spatial distribution and age structure. In the paper, the problem is reduced to two problems of optimal control of ordinary differential equations. The proposed approach is based on the use of exact increment formulas the goal functional. The final result is formulated as a variation optimality condition. An illustrative example is given.



Necessary and sufficient criteria of Lyapunov stability for systems of ordinary differential equations
Abstract
Necessary and sufficient criteria of Lyapunov stability for systems of ordinary differential equations are obtained. The criteria are obtained in the multiplicative form based on the transformation of difference schemes for numerical integration and are converted to the additive and integral forms. The formal restrictions for these criteria are constructed and their applicability conditions are indicated.



Operator methods of the search for extremal controls in linear-quadratic optimal control problems
Abstract
In the class of bilinear control systems with a state-quadratic optimality criterion, new methods of the search for extremal controls are considered. The approach proposed is based on special versions of the maximum principle that have the form of operator fixed-point problems in the space of controls, which are equivalent to the well-known condition of the maximum principle in the class of linear-quadratic optimal control problems. The operator forms of optimality conditions allows one to construct new iterative algorithms for finding controls satisfy the maximum principle. The comparative efficiency of the operator methods is illustrated by numerical simulations of a well-known model optimization problem for a quantum system characterized by degenerate extremal controls.



First boundary-value problem for a class of elliptic systems in a half-space
Abstract
Using the Fourier transform, we examine the first boundary-value problem for two elliptic systems in a half-space. We prove that for both systems, the homogeneous problem has infinitely many solutions depending on one arbitrary function. At the same time, one of the systems is strongly connected under certain conditions for the coefficients of the system, whereas the second system is always strongly connected.



On the exact solution of a certain system of hyperbolic differential equations
Abstract
In this paper, we construct an exact solution of a first-order hyperbolic system of partial differential equations containing linear and quadratic nonlinear equations. Also, we construct a solution of the initial-value (Cauchy) problem for a linear hyperbolic system and a solution of an initial-boundary-value problem for a nonlinear hyperbolic system.



Closed-loop state feedback in linear problems of terminal control
Abstract
We consider the optimal control problem for a linear discrete system with unknown limited disturbances, which must be transferred to a terminal set in a finite time, while providing a minimum guaranteed value of the terminal quality criterion. We discuss two approaches to constructing optimal feedbacks: the disconnectable feedback, which is determined through optimal guarantee programs, and the closed feedback based on optimal control strategies with closures. We discuss disadvantages of the first approach and propose an effective method of constructing optimal closed real-time feedback.



Methods for improving the efficiency of the positional minimum principle in optimal control problems
Abstract
The positional minimum principle is a necessary condition of global optimality, which strengthen the Pontryagin maximum principle and various extremal conditions for smooth and nonsmooth problems. It is based on iterations of the positional descent over the functional related to extremal strategies with respect to a solution of the corresponding Hamilton–Jacobi inequality. We discuss the main methods that allow one to increase the efficiency of positional descent iterations for uncertain extreme strategies and <



On the asymptotics of the Goursat problem with a power boundary layer
Abstract
In this paper, we consider the Goursat problem for a partial differential equation containing a small parameter in the coefficient of the highest derivative. For , the order of the equation does not decrease, but a singularity appears, which has the nature of a power boundary layer. A solution of the singularly perturbed Gaussian problem is constructed in the form of a formal series in powers of the small parameter. The asymptotic nature of the constructed series is proved.



On symmetric boolean functions invariant under the Möbius transform
Abstract
The work is devoted to the study of the class of Boolean functions that are invariant under the Möbius transform. In the first part of the paper, we systematize general information on the Möbius transform and its fixed points. In the second part, we consider a class of symmetric Boolean functions that are invariant under the Möbius transform. The relationship of these functions with columns of the Sierpinski triangle is shown. We propose a method for obtaining masks of all such functions as sums of columns of the Sierpinski triangle. For the case , we proved that a symmetric function is invariant if and only if its mask is invariant.



On some zero-front solutions of an evolution parabolic system
Abstract
We present an existence and uniqueness theorem for a nontrivial analytical zero-front solution of a problem for a nonlinear evolution parabolic predator-prey system. In special cases, we construct exact solutions by reduction to the Cauchy problem for a system of ordinary differential equations, which inherits all features of the original problem. We propose an algorithm for the numerical solution of the problem based on the method of specific solutions and present the result of computational experiments.



Two sequential test schemes with aftereffect
Abstract
Two variants of the urn scheme with aftereffect are described. Using A- and -schemes of sequential tests, we find an explicit distribution of the number of removed balls of a certain color, obtain numerical characteristics of the distribution, and prove limit theorems.



Hierarchical structures and combinatorial problems of information retrieval
Abstract
We examine combinatorial objects of pyramidal structure. We consider one of the ways of representing rules in hierarchical, sequential structures: the method of decision trees, where each object corresponds to a single node that provides a solution. An algorithm for constructing a decision tree based on the generalized Pascal pyramid is suggested. Also, we propose a method for constructing a search index, which displays the proportion of relevant material and allows one to perform comparisons in the variety of terms based on the weight coefficients of terms and paths.






Stabilization of stationary motions of a satellite near the center of mass in a geomagnetic field. V
Abstract
In this paper, we consider problems of stabilization of stationary motions (equilibrium positions and regular precessions) of a satellite near the center of mass in gravitational and magnetic fields under the assumption that the center of mass moves in a circular orbit. Solutions for a number of problems of stabilizing stationary motions of a satellite with the help of magnetic systems are proposed. We present the results of mathematical modeling of the algorithms, which confirm the effectiveness of the developed methodology. This paper is the final part of the work. The first part is: Itogi Nauki i Tekhniki. Sovremennaya Matematika i Ee Prilozheniya. Tematicheskie Obzory. — 2023. — 220. — P. 71–85. The second part is: Itogi Nauki i Tekhniki. Sovremennaya Matematika i Ee Prilozheniya. Tematicheskie Obzory. — 2023. — 221. — P. 71–92. The third part is: Itogi Nauki i Tekhniki. Sovremennaya Matematika i Ee Prilozheniya. Tematicheskie Obzory. — 2023. — 222. — P. 42–63. The fourth part is: Itogi Nauki i Tekhniki. Sovremennaya Matematika i Ee Prilozheniya. Tematicheskie Obzory. — 2023. — 223. — P. 84–106.



On the identification Volterra kernels in integral models of linear nonstationary dynamical systems
Abstract
In this paper, we propose an identification algorithm for a nonstationary linear dynamical system. Conceptually, this algorithm is based on the use of piecewise linear test signals and the reduction of the original problem to a Volterra integral equation of the first kind with two variable integration limits. The numerical implementation of this algorithm is based on the quadrature formula of the middle rectangles and the product integration method. The convergence of the method of middle rectangles for a new class of linear Volterra integral equations is examined.



Variational method for solving a coefficient inverse problem for a parabolic equation with integral conditions
Abstract
In this paper, we consider an inverse control-type problem of determining the minor coefficient of a parabolic equation with an integral boundary-value condition and an additional integral condition. The well-posedness of the problem is examined. The Fréchet differentiability of the target functional based on the additional integral condition is proved and an expression for its gradient is found. A necessary condition for the optimality of control is established.



Projection methods for improving controls in nonlinear control systems with terminal constraints
Abstract
In this paper, we consider a new approach to the optimization of nonlinear control systems with terminal constraints based on the consecutive solution of nonlocal control improvement problems in the form of special systems of functional equations in the control space. The corresponding systems are constructed as fixed-point problems for special control operators with an additional algebraic equation. The methods of successive approximations of the control that preserve terminal constraints at each iteration used in this paper do not contain the time-consuming operation of parametric variation of the control, which is typical for common gradient methods.



Singular systems of differential equations in Banach spaces
Abstract
Degenerate linear systems of differential equations of a special form in Banach spaces are considered. The structure of the solution of the Cauchy problem for such systems is completely determined by the properties of the matrix and operator pencils of the system. Solutions are constructed in the space of distributions with support bounded on the left and are restored using the matrix fundamental operator function of the system. Based on the analysis of the generalized solution constructed in this way, one can obtain theorems on the solvability in the space of functions of finite smoothness of the original Cauchy problem.


