


Vol 304, No Suppl 1 (2019)
- Year: 2019
- Articles: 20
- URL: https://journals.rcsi.science/0081-5438/issue/view/10773
Article
Sergei Vladimirovich Matveev



On an Optimal Control Problem with Discontinuous Integrand
Abstract
We consider an optimal control problem for an autonomous differential inclusion with free terminal time and a mixed functional which contains the characteristic function of a given open set M ⊂ ℝn in the integral term. The statement of the problem weakens the statement of the classical optimal control problem with state constraints to the case where the presence of admissible trajectories of the system in the set M is physically allowed but undesirable due to safety or instability reasons. Using an approximation approach, necessary conditions for the optimality of an admissible trajectory are obtained in the form of Clarke’s Hamiltonian inclusion. The result involves a nonstandard stationarity condition for the Hamiltonian. As in the case of the problem with a state constraint, the obtained necessary optimality conditions may degenerate. Conditions guaranteeing their nondegeneracy and pointwise nontriviality are presented. The results obtained extend the author’s previous results to the case of a problem with free terminal time and more general functional.



The Set of Target Vectors in a Semi-Infinite Linear Program with a Duality Gap
Abstract
We propose a geometric method for the analysis of duality relations in a pair of semi-infinite linear programs (SILPs). The method is based on the use of the conic hull of the coefficients in the constraint system. A relation between the presence of a duality gap and the nonclosedness of the boundary of the conic hull of points in a multidimensional space is established. The geometric approach is used to construct an opposite pair of dual problems and to explore the duality relations for this pair. We construct a nontrivial example of a SILP in which the duality gap occurs for noncollinear target vectors.



Knot Groups and Residual Nilpotence
Abstract
We study groups of classical links, welded links, and virtual links. For classical braids, it is proved that the closures of a braid and its automorphic image are weakly equivalent. This implies the affirmative answer to the question of the coincidence of the groups constructed from a braid and from its automorphic image. We also study the problem of residual nilpotence of groups of virtual knots. It is known that, in a classical knot group, the commutator subgroup coincides with the third term of the lower central series and, hence, the quotient by the terms of the lower central series yields nothing. We prove that the situation is different for virtual knots. A nontrivial homomorphism of the virtual trefoil group to a nilpotent group of class 4 is constructed. We use the Magnus representation of a free group by power series to construct a homomorphism of the virtual trefoil group to a finite-dimensional algebra. This produces the nontrivial linear representation of the virtual trefoil group by unitriangular matrices of order 8.



Optimal Trajectory in ℝ2 under Observation
Abstract
We study the problem of forming a trajectory in a given “corridor” from ℝ2 such that the minimum distance from this trajectory to observers is maximal. Each observer is located outside the corridor and has an open convex observation cone overlapping the corridor. The positions of the observers and the cones are fixed. An observer can measure the distance to an object moving along the trajectory when the object is inside the observer’s cone. We describe an “optimal corridor,” i.e., the set of all optimal trajectories with given initial and terminal points. A similar problem is solved in the case where the moving object is a solid body, more exactly, a disk. For practical calculations, we propose algorithms that construct an optimal corridor and a shortest optimal trajectory for a solid object in a discrete statement. The initial continuous conditions of the problem, such as the boundaries of the corridor and the observation cones, are projected onto a discrete regular grid, and a discrete realization of the optimal corridor and its boundaries are constructed on the grid in the form of 8-connected sequences of grid nodes. The shortest optimal trajectory of the solid object is found with the use of Dijkstra’s algorithm.



Game Problems of Approach for Quasilinear Systems of General Form
Abstract
A conflict-controlled process of the approach of a trajectory to a cylindrical terminal set is studied. The problem statement encompasses a wide range of quasilinear functional-differential systems. We use the technique of set-valued mappings and their selections to derive sufficient conditions for the game termination in a finite time. The methodology used is close to the scheme that involves the time of the first absorption. By way of illustration, quasilinear integro-differential games are examined. For this purpose, their solutions are presented in the form of an analog of the Cauchy formula. The calculations are performed for the case of a system with a simple matrix; the control sets of the players are balls centered at the origin and the terminal set is a linear subspace. Depending on the relations between the initial state of the system and the parameters of the process, sufficient conditions for the game termination are derived. An explicit form of the guaranteed time is found in one specific case.



Automorphisms of an AT4(4, 4, 2)-Graph and of the Corresponding Strongly Regular Graphs
Abstract
Makhnev, Paduchikh, and Khamgokova gave a classification of distance-regular locally GQ(5, 3)-graphs. In particular, there arises an AT 4(4, 4, 2)-graph on 644 vertices with intersection array {96, 75, 16, 1; 1, 16, 75, 96}. The same authors proved that an AT 4(4, 4, 2)-graph is not a locally GQ(5, 3)-graph. However, the existence of an AT 4(4, 4, 2)-graph that is a locally pseudo-GQ(5, 3)-graph is unknown. The antipodal quotient of an AT 4(4, 4, 2)-graph is a strongly regular graph with parameters (322, 96, 20, 32). These two graphs are locally pseudo-GQ(5, 3)-graphs. We find their possible automorphisms. It turns out that the automorphism group of a distance-regular graph with intersection array {96, 75, 16, 1; 1, 16, 75, 96} acts intransitively on the set of its antipodal classes.



Approximation of Minimax Solutions of Hamilton—Jacobi Functional Equations for Time-Delay Systems
Abstract
A minimax solution of the Cauchy problem for a functional Hamilton-Jacobi equation with coinvariant derivatives and a condition at the right end is considered. Hamilton-Jacobi equations of this type arise in dynamical optimization problems for time-delay systems. Their approximation is associated with additional issues of a valid transition from an infinite-dimensional functional argument of the desired solution to a finite-dimensional argument. Earlier, the schemes based on the piecewise linear approximation of the functional argument and the correctness properties of minimax solutions were studied. In this paper, a scheme for the approximation of Hamilton-Jacobi functional equations with coinvariant derivatives by usual Hamilton-Jacobi partial differential equations is proposed and justified. The scheme is based on the approximation of the characteristic functional-differential inclusions used in the definition of the desired minimax solution by ordinary differential inclusions.



On the Geometry of Reachable Sets for Control Systems with Isoperimetric Constraints
Abstract
We consider a nonlinear control system that is linear in the control variables. The control and the trajectory are subject to a system of isoperimetric constraints in the form of inequalities for integral functionals. We describe the boundary of the reachable set of the system at a given time and show that an admissible control taking the system to the boundary of the admissible set is a weakly efficient solution of a certain optimal control problem with a vector criterion if the linearized system is completely controllable. The components of the criterion are integral functionals that specify isoperimetric constraints. The stated result generalizes the authors’ earlier results to the case of several consistent integral constraints. The proof is based on the Graves theorem on covering mappings and on the properties of the derivative of the “input-output” mapping and of the constraints. The result remains valid if the initial state of the system is not fixed but belongs to a given set. The problem is reduced to a control problem with a scalar criterion depending on parameters. The Chebyshev convolution of integral functionals is chosen as the scalar criterion. Necessary conditions are obtained for the optimality of controls taking the system to the boundary of the reachable set in the form of Pontryagin’s maximum principle.



Steiner Problem in the Gromov-Hausdorff Space: The Case of Finite Metric Spaces
Abstract
The Steiner problem is considered in the Gromov-Hausdorff space, i.e., in the space of compact metric spaces (considered up to isometry) endowed with the Gromov-Hausdorff distance. Since this space is not boundedly compact, the problem of the existence of a shortest network in this space is open. It is shown that each finite family of finite metric spaces can be connected by a shortest network. Moreover, it turns out that in this case there exists a shortest tree all of whose vertices are finite metric spaces. An estimate for the number of points in these metric spaces is obtained. As an example, the case of three-point metric spaces is considered. It is also shown that the Gromov-Hausdorff space does not realize minimal fillings; i.e., shortest trees in this space need not be minimal fillings of their boundaries.



Solvability of a Mixed Boundary Value Problem for a Stationary Reaction-Convection-Diffusion Model
Abstract
We study the solvability of an inhomogeneous mixed boundary value problem for a stationary reaction-convection-diffusion model. Such models are often used in science and engineering for the description and analysis of various processes of heat and mass transfer. We focus on the issues of solvability of the boundary value problem in different functional spaces and on the stability of the solution and its continuous dependence on the input data in natural metrics. The peculiarity of the problem consists in the inhomogeneity and irregularity of the mixed boundary data. These boundary data, in general, cannot be continued inside the domain so that the continuation is sufficiently smooth and can be used in the known way to transform the problem to a problem with homogeneous boundary data. To prove the solvability of the problem, we use the Lax-Milgram theorem. Estimates for the norms of the solution follow from the same theorem. The cases where the solution operator is completely continuous are also established. The properties of the solution of the direct problem found in this study will be used later to solve inverse problems.



Automorphisms of a Strongly Regular Graph with Parameters (1305, 440, 115, 165)
Abstract
A graph Γ is called t-isoregular if, for any i ≤ t and any i-vertex subset S, the number Γ(S) depends only on the isomorphism class of the subgraph induced by S. A graph Γ on v vertices is called absolutely isoregular if it is (v — 1)-isoregular. It is known that each 5-isoregular graph is absolutely isoregular, and such graphs have been fully described. Each precisely 4-isoregular graph is either a pseudogeometric graph for pGr(2r, 2r3 + 3r2 — 1) or its complement. A pseudogeometric graph for pGr(2r, 2r3 + 3r2 — 1) is denoted by Izo(r). Graphs Izo(r) do not exist for an infinite set of values of r (r = 3, 4, 6, 10,…). The existence of Izo(5) is unknown. In this work, we find possible automorphisms for the neighborhood of an edge from Izo(5).



On the Problem of Input Reconstruction in a Nonlinear System with Constant Delay
Abstract
We study the problem of reconstructing an unknown input acting on a system described by a nonlinear vector differential equation with constant delay. Both the input and the solution (trajectory) of the system are unknown. During the operation of the system, its phase states are measured at discrete times. The measurements, in general, are inaccurate. It is required to give a dynamic stable rule for the approximate reconstruction of the input, which means that the approximate values must be found in real time and the approximations must be arbitrarily accurate for sufficiently exact observations. For the solution of this problem, we propose an algorithm based on the method of models with feedback control. The algorithm reconstructs the unknown input simultaneously with the process. The algorithm is stable with respect to information noises and computational errors.



On the Oikawa and Arakawa Theorems for Graphs
Abstract
The present paper is devoted to the further development of the discrete theory of Riemann surfaces, which was started in the papers by M. Baker and S. Norine and their followers at the beginning of the century. This theory considers finite graphs as analogs of Riemann surfaces and branched coverings of graphs as holomorphic mappings. The genus of a graph is defined as the rank of its fundamental group. The main object of investigation in the paper is automorphism groups of a graph acting freely on the set of half-edges of the graph. These groups are discrete analogs of groups of conformal automorphisms of a Riemann surface. The celebrated Hurwitz theorem (1893) states that the order of the group of conformal automorphisms of a compact Riemann surface of genus g > 1 does not exceed 84(g — 1). Later, K. Oikawa and T. Arakawa refined this bound in the case of groups that fix several finite sets of prescribed cardinalities. This paper provides proofs of discrete versions of the mentioned theorems. In addition, a discrete analog of the E. Bujalance and G. Gromadzki theorem improving one of Arakawa’s results is obtained.



A Metanilpotency Criterion for a Finite Solvable Group
Abstract
Denote by |x| the order of an element x of a group. An element of a group is called primary if its order is a nonnegative integer power of a prime. If a and b are primary elements of coprime orders of a group, then the commutator a−1b−1ab is called a *-commutator. The intersection of all normal subgroups of a group such that the quotient groups by them are nilpotent is called the nilpotent residual of the group. It is established that the nilpotent residual of a finite group is generated by commutators of primary elements of coprime orders. It is proved that the nilpotent residual of a finite solvable group is nilpotent if and only if |ab| ≥ |a||b| for any *-commutators of a and b of coprime orders.



On Asymptotic Properties of Solutions of Control Systems with Random Parameters
Abstract
Differential equations and control systems with impulse action and random parameters are considered. These objects are characterized by stochastic behavior: the lengths θk of the intervals between the times of the impulses τk, k = 0,1,…, are random variables and the magnitudes of the impulses also depend on random actions. The basic object of research is the control system



Virtual 3-Manifolds of Complexity 1 and 2
Abstract
Matveev in 2009 introduced the notion of virtual 3-manifold, which generalizes the classical notion of 3-manifold. A virtual 3-manifold is an equivalence class of so-called special polyhedra. Each virtual 3-manifold determines a 3-manifold with nonempty boundary and ℝP2-singularities. Many invariants of manifolds, such as Turaev–Viro invariants, can be extended to virtual 3-manifolds. The complexity of a virtual 3-manifold is k if its equivalence class contains a special polyhedron with k true vertices and contains no special polyhedra with fewer true vertices. In this paper, we give a complete list of virtual 3-manifolds of complexity 1 and present two-sided bounds for the number of virtual 3-manifolds of complexity 2. The question of the complete classification for virtual 3-manifolds of complexity 2 remains open.



Impulse Differential Game with a Mixed Constraint on the Choice of the Control of the First Player
Abstract
We consider a linear differential game in which the first player can choose both an impulse control and a control subject to a geometric constraint. The first player can use a prescribed amount of resource to form the impulse control. Portions of this amount can be separated at certain times, thus producing “instantaneous” changes of the state vector and complicating the problem. The control of the second player is subject to a geometric constraint. The vectograms of the players are described by the same ball with different time-dependent radii. The terminal set is a ball with fixed radius. The aim of the first player is to bring the state vector to the terminal set at a given time. The aim of the second player is opposite. Necessary and sufficient conditions for meeting the terminal set at the given time are found, and the corresponding controls of the players guaranteeing the achievement of their goals are constructed. A solution of an example illustrating the theory is given.



Brieskorn Manifolds, Generalized Sieradski Groups, and Coverings of Lens Spaces
Abstract
A Brieskorn manifold B(p, q, r) is the r-fold cyclic covering of the 3-sphere S3 branched over the torus knot T(p, q). Generalized Sieradski groups S(m, p, q) are groups with an m-cyclic presentation Gm(w), where the word w has a special form depending on p and q. In particular, S(m, 3, 2) = Gm(w) is the group with m generators x1,…, xm and m defining relations w(xi, xi+1, xi+2) = 1, where w(xi, xi+1, xi+2) = xi, xi+2, xi+1−1. Cyclic presentations of the groups S(2n, 3, 2) in the form Gn(w) were investigated by Howie and Williams, who showed that the n-cyclic presentations are geometric, i.e., correspond to spines of closed 3-manifolds. We establish a similar result for the groups S(2n, 5, 2). It is shown that in both cases the manifolds are n-fold branched cyclic coverings of lens spaces. To classify some of the constructed manifolds, we use Matveev’s computer program “Recognizer.”



On Finite Simple Linear and Unitary Groups over Fields of Different Characteristics with Coinciding Prime Graphs. I
Abstract
Suppose that G is a finite group, π(G) is the set of prime divisors of its order, and ω(G) is the set of orders of its elements. We define a graph on π(G) with the following adjacency relation: different vertices r and s from π(G) are adjacent if and only if rs ∈ ω(G). This graph is called the Gruenberg-Kegel graph or the prime graph of G and is denoted by GK(G). In a series of papers, we describe the coincidence conditions for the prime graphs of nonisomorphic simple groups. This issue is connected with Vasil’ev’s Question 16.26 in the Kourovka Notebook about the number of nonisomorphic simple groups with the same prime graph. Earlier the author derived necessary and sufficient conditions for the coincidence of the prime graphs of two nonisomorphic finite simple groups of Lie type over fields of orders q and q1, respectively, with the same characteristic. Let G and G1 be two nonisomorphic finite simple groups of Lie type over fields of orders q and q1, respectively, with different characteristics. The author also obtained necessary conditions for the coincidence of the prime graphs of two nonisomorphic finite simple groups of Lie type. In the present paper the latter result is refined in the case where G is a simple linear group of sufficiently high Lie rank over a field of order q. If G is a simple linear group of sufficiently high Lie rank, then we prove that the prime graphs of G and G1 may coincide only in one of the nineteen cases. As corollaries of the main result, we obtain constraints (under some additional conditions) on the possible number of simple groups whose prime graph is the same as the prime graph of a simple linear group.


