On saturation effects in the Neumann boundary control of elliptic optimal control problems

A Neumann boundary control problem for a linear-quadratic elliptic optimal control problem in a polygonal domain is investigated. The main goal is to show an optimal approximation order for discretized problems after a postprocessing process. It turns out that two saturation processes occur: The regularity of the boundary data of the adjoint state is limited if the largest angle of the polygon is at least 2π /3. Moreover, piecewise linear finite elements cannot guarantee the optimal order, if the largest angle of the polygon is greater than π /2. We will derive error estimates of order h^σ with σ ∈ [3/2, 2] depending on the largest angle and properties of the finite elements. (© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim)     

Large dimensional sets not containing a given angle

We say that a set in a Euclidean space does not contain an angle α if the angle determined by any three points of the set is not equal to α. The goal of this paper is to construct compact sets of large Hausdorff dimension that do not contain a given angle α ∈ (0,π). We will construct such sets in ℝ ⁿ of Hausdorff dimension c(α)n with a positive c(α) depending only on α provided that α is different from π/3, π/2 and 2π/3. This improves on an earlier construction (due to several authors) that has dimension c(α) log n.     

Virtual control regularization of state constrained linear quadratic optimal control problems

A numerical method for linear quadratic optimal control problems with pure state constraints is analyzed. Using the virtual control concept introduced by Cherednichenko et al. (Inverse Probl. 24:1–21, 2008) and Krumbiegel and R?sch (Control Cybern. 37(2):369–392, 2008), the state constrained optimal control problem is embedded into a family of optimal control problems with mixed control-state constraints using a regularization parameter α>0. It is shown that the solutions of the problems with mixed control-state constraints converge to the solution of the state constrained problem in the L ² norm as α tends to zero. The regularized problems can be solved by a semi-smooth Newton method for every α>0 and thus the solution of the original state constrained problem can be approximated arbitrarily close as α approaches zero. Two numerical examples with benchmark problems are provided.     

Localization near the corner point of the principal eigenfunction of the Dirichlet problem in a domain with thin edging

We establish that the principal eigenfunction of the Dirichlet problem in a domain with a thin heavy edging admits localization near the corner point of opening angle α > π. The edging amounts to a boundary strip of small width ɛ with the density function ɛ ^−2−m , m > 0, while it is O(1) in the remaining part of the domain. We derive the result by analyzing the essential and discrete spectra of an auxiliary problem in an infinite angle without the small parameter. We state several open questions about the structure of spectra of both problems.     

An Isoperimetric Theorem in Plane Geometry

Abstract. Let P be a simple polygon. Let the vertices of P be mapped, according to a counterclockwise traversal of the boundary, into a strictly increasing sequence of real numbers in [0, 2π) . Let a ray be drawn from each vertex so that the angle formed by the ray and a horizontal line pointing to the right equals, in measure, the number mapped to the vertex. Whenever the rays from two consecutive vertices intersect, let them induce the triangular region with extreme points comprising the vertices and the intersection point. It is shown that there is a fixed α such that if all of the assigned angles are increased by α , the triangular regions induced by the redirected rays cover the interior of P . This covering implies the standard isoperimetric inequalities in two dimensions, as well as several new inequalities, and resolves a question posed by Yaglom and Boltanskii.     

A boundary-value problem for the biharmonic equation and the iterated Laplacian in a 3D-domain with an edge

Let Ω be a domain with piecewise smooth boundary. In general, it is impossible to obtain a generalized solution u ∈ W ₂ ² (Ω) of the equation Δ _x ² u = f with the boundary conditions u = Δ_xu = 0 by solving iteratively a system of two Poisson equations under homogeneous Dirichlet conditions. Such a system is obtained by setting v = −Δu. In the two-dimensional case, this fact is known as the Sapongyan paradox in the theory of simply supported polygonal plates. In the present paper, the three-dimensional problem is investigated for a domain with a smooth edge Γ. If the variable opening angle α ∈ C^∞(Γ) is less than π everywhere on the edge, then the boundary-value problem for the biharmonic equation is equivalent to the iterated Dirichlet problem, and its solution u inherits the positivity preserving property from these problems. In the case α ∈ (π 2π), the procedure of solving the two Dirichlet problems must be modified by permitting infinite-dimensional kernel and co-kernel of the operators and determining the solution u ∈ W ₂ ² (Ω) by inverting a certain integral operator on the contour Γ. If α(s) ∈ (3π/2,2π) for a point s ∈ Γ, then there exists a nonnegative function f ∈ L₂(Ω) for which the solution u changes sign inside the domain Ω. In the case of crack (α = 2π everywhere on Γ), one needs to introduce a special scale of weighted function spaces. In this case, the positivity preserving property fails. In some geometrical situations, the problems on well-posedness for the boundary-value problem for the biharmonic equation and the positivity property remain open. Bibliography: 46 titles. __________ Translated from Zapiski Nauchnykh Seminarov POMI, Vol. 336, 2006, pp. 153–198.     

An Isoperimetric Theorem in Plane Geometry

   Abstract. Let P be a simple polygon. Let the vertices of P be mapped, according to a counterclockwise traversal of the boundary, into a strictly increasing sequence of real numbers in [0, 2π) . Let a ray be drawn from each vertex so that the angle formed by the ray and a horizontal line pointing to the right equals, in measure, the number mapped to the vertex. Whenever the rays from two consecutive vertices intersect, let them induce the triangular region with extreme points comprising the vertices and the intersection point. It is shown that there is a fixed α such that if all of the assigned angles are increased by α , the triangular regions induced by the redirected rays cover the interior of P . This covering implies the standard isoperimetric inequalities in two dimensions, as well as several new inequalities, and resolves a question posed by Yaglom and Boltanskii.     

Asphericity of shadows of a convex body

A shadow F of a body K is a parallel projection of K to a plane. The shadow F is said to be ε-aspherical if the boundary ∂F lies in a circular ring with center O and ratio of radii equal to 1 + ε. F is said to be ε-aspherical by a part of α if the same is true for the part of ∂F lying inside an angle of 2α π with vertex at O (or within the union of two vertical angles equal to απ if K is centrally symmetric). It is proved that each convex body K ⊂ ℝ³ has a -aspherical shadow and a shadow that is (sec π/5 − 1)-aspherical by 4/5. If K is centrally symmetric, then K has a -aspherical shadow and a shadow that is (sec π/7 − 1)-aspherical by 6/7. Bibliography: 5 titles. __________ Translated from Zapiski Nauchnykh Seminarov POMI, Vol. 329, 2005, pp. 67–78.     

The Union of Congruent Cubes in Three Dimensions

Abstract. A dihedral (trihedral) wedge is the intersection of two (resp. three) half-spaces in R ³ . It is called α-fat if the angle (resp., solid angle) determined by these half-spaces is at least α>0 . If, in addition, the sum of the three face angles of a trihedral wedge is at least γ >4π/3 , then it is called (γ,α)-substantially fat . We prove that, for any fixed γ>4π/3, α>0 , the combinatorial complexity of the union of n (a) α -fat dihedral wedges, and (b) (γ,α) -substantially fat trihedral wedges is at most O(n^ 2+ ɛ ) , for any ɛ >0 , where the constants of proportionality depend on ɛ , α (and γ ). We obtain as a corollary that the same upper bound holds for the combinatorial complexity of the union of n (nearly) congruent cubes in R ³ . These bounds are not far from being optimal.     

The Union of Congruent Cubes in Three Dimensions

   Abstract. A dihedral (trihedral) wedge is the intersection of two (resp. three) half-spaces in R ³ . It is called α-fat if the angle (resp., solid angle) determined by these half-spaces is at least α>0 . If, in addition, the sum of the three face angles of a trihedral wedge is at least γ >4π/3 , then it is called (γ,α)-substantially fat . We prove that, for any fixed γ>4π/3, α>0 , the combinatorial complexity of the union of n (a) α -fat dihedral wedges, and (b) (γ,α) -substantially fat trihedral wedges is at most O(n^ 2+ ɛ ) , for any ɛ >0 , where the constants of proportionality depend on ɛ , α (and γ ). We obtain as a corollary that the same upper bound holds for the combinatorial complexity of the union of n (nearly) congruent cubes in R ³ . These bounds are not far from being optimal.     

Block separation in solvable groups

If π is a set of primes, a finite group G is block π-separated if for every two distinct irreducible complex characters α, β ∈ Irr(G) there exists a prime p ∈ π such that α and β lie in different Brauer p-blocks. A group G is block separated if it is separated by the set of prime divisors of |G|. Given a set π with n different primes, we construct an example of a solvable π-group G which is block separated but it is not separated by every proper subset of π. Received: 22 December 2004     

The maximum angle condition is not necessary for convergence of the finite element method

We show that the famous maximum angle condition in the finite element analysis is not necessary to achieve the optimal convergence rate when simplicial finite elements are used to solve elliptic problems. This condition is only sufficient. In fact, finite element approximations may converge even though some dihedral angles of simplicial elements tend to π.     

On a relation between the Sylow and Baer-Suzuki theorems

Given a set π of primes, say that a finite group G satisfies the Sylow π-theorem if every two maximal π-subgroups of G are conjugate; equivalently, the full analog of the Sylow theorem holds for π-subgroups. Say also that a finite group G satisfies the Baer-Suzuki π-theorem if every conjugacy class of G every pair of whose elements generate a π-subgroup itself generates a π-subgroup. In this article we prove, using the classification of finite simple groups, that if a finite group satisfies the Sylow π-theorem then it satisfies the Baer-Suzuki π-theorem as well.     

Error estimates for the discretization of elliptic control problems with pointwise control and state constraints

A family of elliptic optimal control problems with pointwise constraints on control and state is considered. We are interested in approximation of the optimal solution by a finite element discretization of the involved partial differential equations. The discretization error for a problem with mixed state constraints is estimated in the semidiscrete case and in the fully discrete scheme with the convergence of order h|ln h| and h ^1/2, respectively. However, considering the unregularized continuous problem and the discrete regularized version, and choosing suitable relation between the regularization parameter and the mesh size, i.e., ε∼h ², a convergence order arbitrary close to 1, i.e., h ^1−β is obtained. Therefore, we benefit from tuning the involved parameters.     

A Priori Error Estimates for Finite Element Methods for H (2,1)-Elliptic Equations

The convergence of finite element methods for linear elliptic boundary value problems of second and forth order is well understood. In this article, we introduce finite element approximations of some linear semi-elliptic boundary value problem of mixed order on a two-dimensional rectangular domain Q. The equation is of second order in one direction and forth order in the other and appears in the optimal control of parabolic partial differential equations if one eliminates the control and the state (or the adjoint state) in the first order optimality conditions. We establish a regularity result and estimate for the finite element error of conforming approximations of this equation. The finite elements in use have a tensor product structure, in one dimension we use linear, quadratic or cubic Lagrange elements in the other dimension cubic Hermite elements. For these elements, we prove the error bound O(h ² + τ ^k ) in the energy norm and O((h ² + τ ^k )(h ² + τ)) in the L ²(Q)-norm.     

Uniquely factorizable elements and solvability of finite groups

In a finite group G every element can be factorized in such a way that there is one factor for each prime divisor p of | G |, and the order of this factor is p^α for some integer α ≧ 0. We define g ∈G to be uniquely factorizable if it has just one such factorization (whose factors must be pairwise commuting). We consider the existence of uniquely factorizable elements and its relation to the solvability of the group. We prove that G is solvable if and only if the set of all uniquely factorizable elements of G is the Fitting subgroup of G. We also prove various sufficient conditions for the non-existence of uniquely factorizable elements in non-solvable groups. Received: 9 June 2005     

Convergence of Distributed Optimal Controls on the Internal Energy in Mixed Elliptic Problems when the Heat Transfer Coefficient Goes to Infinity

We consider a steady-state heat conduction problem P _α with mixed boundary conditions for the Poisson equation depending on a positive parameter α , which represents the heat transfer coefficient on a portion Γ ₁ of the boundary of a given bounded domain in R ⁿ . We formulate distributed optimal control problems over the internal energy g for each α . We prove that the optimal control g_ op _α and its corresponding system u_ g_ op _α α and adjoint p_ g_ op _α α states for each α are strongly convergent to g _op , u_ g _op and p _ g _op , respectively, in adequate functional spaces. We also prove that these limit functions are respectively the optimal control, and the system and adjoint states corresponding to another distributed optimal control problem for the same Poisson equation with a different boundary condition on the portion Γ ₁ . We use the fixed point and elliptic variational inequality theories.     

Convergence of Distributed Optimal Controls on the Internal Energy in Mixed Elliptic Problems when the Heat Transfer Coefficient Goes to Infinity

We consider a steady-state heat conduction problem P _α with mixed boundary conditions for the Poisson equation depending on a positive parameter α , which represents the heat transfer coefficient on a portion Γ ₁ of the boundary of a given bounded domain in R ⁿ . We formulate distributed optimal control problems over the internal energy g for each α . We prove that the optimal control g_ op _α and its corresponding system u_ g_ op _α α and adjoint p_ g_ op _α α states for each α are strongly convergent to g _op , u_ g _op and p _ g _op , respectively, in adequate functional spaces. We also prove that these limit functions are respectively the optimal control, and the system and adjoint states corresponding to another distributed optimal control problem for the same Poisson equation with a different boundary condition on the portion Γ ₁ . We use the fixed point and elliptic variational inequality theories.     

When Does the Free Boundary Enter into Corner Points of the Fixed Boundary?

Our prime goal in this note is to lay the ground for studying free boundaries close to the corner points of a fixed Lipschitz continuous boundary. Our study is restricted to 2-space dimensions and to the obstacle problem. Our main result states that the free boundary cannot enter a corner x⁰ of the fixed boundary if the (interior) angle is less than π, provided that the boundary datum is zero near to the point x⁰. For larger angles and other boundary data, the free boundary may enter into corners, as discussed in the text. Bibliography: 10 titles. To Nina Nikolaevna Uraltseva on the occasion of her 70th birthday __________ Published in Zapiski Nauchnykh Seminarov POMI, Vol. 310, 2004, pp. 213–225.