Operator Ranges in Banach Spaces,I

Let X be a Banach space. A subspace L of X is called an operator range if there exists a continuous linear operator T defined on some Banach space Y and such that TY = L. If Y = X then L is called an endomorphism range. The paper investigates operator ranges under the following perspectives: (1) Existence (Section 3), (2) Inclusion (Section 4), and (3) Decomposition (Section 5). Section 3 considers the existence in X of operator ranges satisfying certain conditions. The main result is the following: if X and Fare separable Banach spaces and T : Y → X is a continuous operator with nonclosed range, then there exists a nuclear operator R:Y→X such that T + R is injective and has nonclosed dense range (Theorem 3.2). Section 4 seeks to determine conditions under which every nonclosed operator range in a Banach space is contained in the range of some injective endomorphism with nonclosed dense range. Theorem 4.3 contains a sufficient condition for this. Examples of spaces satisfying this condition are c₀, l_p (1 < p < ∞), L_q (1 < q < 2) and their quotients. In particular, this answers a question posed by W. E. Longstaff and P. Rosenthal (Integral Equations and Operator Theory 9 , (1986), 820-830. Section 5 discusses the possibility of representing a given dense nonclosed operator range as the sum of a pair L_1, L₂ of operator ranges with zero intersection in the cases where (a) L₁ and L₂ are dense, (b) L₁ and L₂ are closed. The results generalize corresponding results, for endomorphisms in Hilbert space, of J. Dixmier (Bull. Soc. Math. France 77 (1949), 11-101 and P. A. Fillmore and J. P. Williams (Adv. Math. 7 (1971), 254-281. A final section is devoted to open problems.     

Evolution Problems involving non-stationary Operators between two Banach Spaces I. Existence and Uniqueness Theorems

Synopsis

(for ‘Evolution Problems involving non-stationary Operators between two Banach Spaces I-II)

In this series of two papers the initial-value problem [B(t)u(t)' = A(t)u(t), Bu(0) = y, with A = A(t) and B = B(t) time-varying operators from one Banach space X to another Banach space Y, and y an arbitrary element of Y, is considered. By making use of the theory of B-evolutions and by integrating certain temporally inhomogeneous equations, a unique solution is obtained for any y in Y. The solution is formulated explicitly in terms of a certain solution operator which involves the B(t)-evolution generated by the closed pair >A(t),B(t)< of operators. Certain properties of the solution operator are also studied. The well-known results, obtained by making use of semigroup theory, for the evolution problem [u(t)]' = A(t)u(t), u(0) = u₀, where A is a closed operator in a Banach space with dense domain, may also be derived from our results.     

Identification of the Time Derivative Coefficient in a Fast Diffusion Degenerate Equation

In this paper, we deal with the identification of the space variable time derivative coefficient u in a degenerate fast diffusion differential inclusion. The function u is vanishing on a subset strictly included in the space domain Ω. This problem is approached as a control problem (P) with the control u. An approximating control problem (P _ε ) is introduced and the existence of an optimal pair is proved. Under certain assumptions on the initial data, the control is found in W ^2,m (Ω), with m>N, in an implicit variational form. Next, it is shown that a sequence of optimal pairs (u_e^*,y_e^*)(u_{\varepsilon }^{\ast },y_{\varepsilon }^{\ast }) of (P _ε ) converges as ε goes to 0 to a pair (u ^*,y ^*) which realizes the minimum in (P), and y ^* is the solution to the original state system.     

Mixing of Cartesian squares of positive operators

LetT be a power bounded positive operator inL ₁(X, Σ, m)of a probability space, given by a transition measureP (x, A). The Cartesian squareS is the operator onL ₁ (X × X, Σ × Σ, m × m) induced by the transition measure Q((x, y), A × B)=P(x, A)P(y, B).T iscompletely mixing if ∝u e dm=0 impliesT ⁿ u→0 weakly (where 0≦e ∈L _∞ withT ^* e=e).Theorem. IfT has no fixed points, thenT is completely mixing if and only ifS is completely mixing. Part of this research was done at The Hebrew University of Jerusalem.     

On the Uniform Poincaré Inequality

We give a proof of the Poincaré inequality in W ^{1, p} (Ω) with a constant that is independent of Ω ? , where  is a set of uniformly bounded and uniformly Lipschitz domains in ? ⁿ . As a byproduct, we obtain the following: The first non vanishing eigenvalues λ₂(Ω) of the standard Neumann (variational) boundary value problem on Ω for the Laplace operator are bounded below by a positive constant if the domains Ω vary and remain uniformly bounded and uniformly Lipschitz regular.     

A globally and superlinearly convergent quasi-Newton method for general box constrained variational inequalities without smoothing approximation

A new quasi-Newton algorithm for the solution of general box constrained variational inequality problem (GVI(l, u, F, f)) is proposed in this paper. It is based on a reformulation of the variational inequality problem as a nonsmooth system of equations by using the median operator. Without smoothing approximation, the proposed quasi-Newton algorithm is directly applied to solve this class of nonsmooth equations. Under appropriate assumptions, it is proved that the algorithmic sequence globally and superlinearly converges to a solution of the equation reformulation and also of GVI(l, u, F, f). Numerical results show that our new algorithm works quite well.     

Reconstruction of a Pair Integral Operator of the Convolution Type

For an arbitrary operator, we pose a general reconstruction problem inverse to the problem of finding solutions. For the pair operator considered, this problem is reduced to the equivalent problem of reconstruction of the kernels of the pair integral equation of the convolution type that generates this operator. In the cases investigated, we prove theorems that characterize the reconstruction of the corresponding kernels, which are constructed in terms of two functions from different Banach algebras of the type L ₁(–, ) with weight.     

Nonlinear Elliptic Variational Inequalities

We consider the problem of finding a solution to a class of nonlinear elliptic variational inequalities. These inequalities may be defined on bounded or unbounded domains Ω, and the nonlinearity can depend on gradient terms. Appropriate definitions of sub-and supersolutions relative to the constraint sets are given. By using a mixture of maximal monotone operator theory and compactness arguments we prove the existence of a H²(Ω) solution lying between a given subsolution φ₁ and a given supersolution φ₂≧φ₁, when Ω is bounded, and a H¹(Ω) solution when Ω is unbounded.     

Orthogonal random vectors and the Hurwitz-Radon-Eckmann theorem

In several different aspects of algebra and topology the following problem is of interest: find the maximal number of unitary antisymmetric operatorsU _i inH = ℝ ⁿ with the propertyU _i U _j = −U _j U _i (i≠j). The solution of this problem is given by the Hurwitz-Radon-Eckmann formula. We generalize this formula in two directions: all the operatorsU _i must commute with a given arbitrary self-adjoint operator andH can be infinite-dimensional. Our second main result deals with the conditions for almost sure orthogonality of two random vectors taking values in a finite or infinite-dimensional Hilbert spaceH. Finally, both results are used to get the formula for the maximal number of pairwise almost surely orthogonal random vectors inH with the same covariance operator and each pair having a linear support inH⊕H. The paper is based on the results obtained jointly with N.P. Kandelaki (see [1,2,3]).     

Multivariational Inequalities and Operator Inclusions in Banach Spaces with Mappings of the Class (S)+

We prove theorems on the existence of solutions of variational inequalities and operator inclusions in Banach spaces with multivalued mappings of the class (S)₊. We justify the method of penalty operators for variational inequalities.     

On the existence of variational principles for a differential-difference evolution operator

For a differential-difference evolution operator, we obtain necessary and sufficient conditions for the existence of a variational principle. We describe the structure of the operators P _λ and Q for which the corresponding differential-difference equation admits a direct variational statement.     

Perturbation of rarefaction waves

Consider a pair of genuinely nonlinear strictly hyperbolic conservation lawsU _t +F(U) _x =0 with initial dataU(O,X)=U _o (X). Suppose that the initial dataU _o (X)=U ₁ (X)+U ₂ (X), whereU ₁ (X) will issue rarefaction waves only,U ₂ (X) has any finite total variation and sufficiently small deviation. We prove that the Cauchy problem has a global solution. This work is supported in part by the Foudation of Zhongshan University Advanced Research Centre.     

Divergence L 2-Coercivity Inequalities

This paper derives sharp L ²-coercivity inequalities for the divergence operator on bounded Lipschitz regions in ? ⁿ . They hold for fields in H(div,Ω) that are orthogonal to N(div). The optimal constants in the inequality are defined by a variational principle and are identified as the least eigenvalue of a nonstandard boundary value problem for a linear biharmonic type operator. The dependence of the optimal constant under dilations of the region is described and a generalization that involves weighted surface integrals is also proved. When n = 2, this also yields a similar coercivity result for the curl operator.     

Multi-Duality in Minimal Surface—Type Problems

The multi-duality of the nonlinear variational problem inf J(u, Λu) is studied for minimal surfaces-type problems. By using the method developed by Gao and Strang [1], the Fenchel-Rockafellar's duality theory is generalized to the problems with affine operator Λ. Two dual variational principles are established for nonparametric surfaces with constant mean curvature. We show that for the same primal problem, there may exist different dual problems. The primal problem may or may not possess a solution, whereas each dual problem possesses a unique solution. An evolutionary method for solving the nonlinear optimal-shape design problem is presented with numerical results.     

Multiple solutions to a class of inclusion problem with the p(x)-Laplacian

In this article, we obtain the existence of at least two nontrivial solutions for a nonlinear elliptic problem involving p(x)-Laplacian type operator and nonsmooth potentials. Our approach is variational and it is based on the nonsmooth critical point theory for locally Lipschitz functions.     

Finding multiple solutions to elliptic systems with polynomial nonlinearity

Elliptic systems with polynomial nonlinearity usually possess multiple solutions. In order to find multiple solutions, such elliptic systems are discretized by eigenfunction expansion method (EEM). Error analysis of the discretization is presented, which is different from the error analysis of EEM for scalar elliptic equations in three aspects: first, the choice of framework for the nonlinear operator and the corresponding isomorphism of the linearized operator; second, the definition of an auxiliary problem in deriving the relation between the L² norm and H¹ norm of the Ritz projection error; third, the bilinearity/nonbilinearity of the linearized variational forms. The symmetric homotopy for the discretized equations preserves not only D₄ symmetry, but also structural symmetry. With the symmetric homotopy, a filter strategy and a finite element Newton refinement, multiple solutions to a system of semilinear elliptic equations arising from Bose–Einstein condensate are found.     

Version of the Grothendieck theorem for subspaces of analytic functions in lattices

A version of Grothendieck’s inequality says that any bounded linear operator acting from a Banach lattice X to a Banach lattice Y acts from X(ℓ²) to Y (ℓ²) as well. A similar statement is proved for Hardy-type subspaces in lattices of measurable functions. Namely, let X be a Banach lattice of measurable functions on the circle, and let an operator T act from the corresponding subspace of analytic functions X_A to a Banach lattice Y or, if Y is also a lattice of measurable functions on the circle, to the quotient space Y/Y_A. Under certain mild conditions on the lattices involved, it is proved that T induces an operator acting from X_A(ℓ²) to Y (ℓ²) or to Y/Y_A(ℓ2), respectively. Bibliography: 7 titles. __________ Translated from Zapiski Nauchnykh Seminarov POMI, Vol. 327, 2005, pp. 5–16.     

Operators with Wentzell boundary conditions and the Dirichlet‐to‐Neumann operator

In this paper we relate the generator property of an operator A with (abstract) generalized Wentzell boundary conditions on a Banach space X and its associated (abstract) Dirichlet‐to‐Neumann operator N acting on a “boundary” space . Our approach is based on similarity transformations and perturbation arguments and allows to split A into an operator A₀₀ with Dirichlet‐type boundary conditions on a space X₀ of states having “zero trace” and the operator N. If A₀₀ generates an analytic semigroup, we obtain under a weak Hille–Yosida type condition that A generates an analytic semigroup on X if and only if N does so on . Here we assume that the (abstract) “trace” operator is bounded that is typically satisfied if X is a space of continuous functions. Concrete applications are made to various second order differential operators.     

Optimal feedback control for constrained mechanical systems

An approach to minimize the control costs and ensuring a stable deviation control is the Riccati controller and we want to use it to control constrained dynamical systems (differential algebraic equations of Index 3). To describe their discrete dynamics, a constrained variational integrators [1] is used. Using a discrete version of the Lagrange-d’Alembert principle yields a forced constrained discrete Euler-Lagrange equation in a position-momentum form that depends on the current and future time steps [2]. The desired optimal trajectory (q_opt, p_opt) and according control input u_opt is determined solving the discrete mechanics and optimal control (DMOC) algorithm [3] based on the variational integrator. Then, during time stepping of the perturbed system, the discrete Riccati equation yields the optimal deviation control input u_R. Adding u_opt and u_R to the discrete Euler-Lagrange equation causes a structure preserving trajectory as both DMOC and Riccati equations are based on the same variational integrator. Furthermore, coordinate transformations are implemented (minimal, redundant and nullspace) enabling the choice of different coordinates in the feedback loop and in the optimal control problem. (© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim)     

Weighted simultaneous Chebyshev approximation

In this paper we discuss the problem of weighted simultaneous Chebyshev approximation to functions f₁,…f_m ε C(X) (1 m ∞), i.e., we wish to minimize the expression {∑^m_{j = 1} λ_j¦f_j − q¦^p}^1/p∞, where λ_j > 0, ∑^m_{j = 1} λ_j = 1, p 1. For this problem we establish the main theorems of the Chebyshev theory, which include the theorems of existence, alternation, de La Vallée Poussin, uniqueness, strong uniqueness, as well as that of continuity of the best approximation operator, etc.