0000000000048358
AUTHOR
Uldis Raitums
Constrained control of a nonlinear two point boundary value problem, I
In this paper we consider an optimal control problem for a nonlinear second order ordinary differential equation with integral constraints. A necessary optimality condition in form of the Pontryagin minimum principle is derived. The proof is based on McShane-variations of the optimal control, a thorough study of their behaviour in dependence of some denning parameters, a generalized Green formula for second order ordinary differential equations with measurable coefficients and certain tools of convex analysis.
Extremal Problems for Elliptic Systems
The specific properties of optimal control problems for elliptic systems, if compared with the case of a single equation, are described. Within them are: strong closures of sets of feasible states; the relaxability via convexification; the type of necessary optimality conditions.
On the strong closure of strains and stresses in linear elasticity
We consider the following special problem related to the optimal layout problems of materials: given two linear elastic materials, the elasticity tensors of which are C1 and C2, and a force f, find the strong closure of strains and stresses as the distribution of the materials varies, or, alternatively, find the sets of elasticity tensors which generate these strong closures. In this paper, it is shown that the local incompatibility conditions depending on C1, C2 and the local properties of strains or stresses completely characterize these sets. A connection to multiple-well problems is established.
Relaxation of a weakly discontinuous functional depending on one control function
The paper considers an optimal control problem of the typewhere the set M of admissible controls consists of all measurable vector‐functions h, which can take only two values h1 or h2. It is shown that the relaxation of this problem can be explicitly computed by rank‐one laminates.
Strictly convergent algorithm for an elliptic equation with nonlocal and nonlinear boundary conditions
The paper describes a formally strictly convergent algorithm for solving a class of elliptic problems with nonlinear and nonlocal boundary conditions, which arise in modeling of the steady-state conductive-radiative heat transfer processes. The proposed algorithm has two levels of iterations, where inner iterations by means of the damped Newton method solve an appropriate elliptic problem with nonlinear, but local boundary conditions, and outer iterations deal with nonlocal terms in boundary conditions.
On dependence of sets of functions on the mean value of their elements
The paper considers, for a given closed bounded set M ⊂ R m and K = (0,1) n ⊂ R n , the set M = {h ϵ L2 (K;R m ) | h(x) ϵ M a.e.x ϵ K} and its subsets It is shown that, if a sequence {hk } ⊂ coM converges to an element hk ϵ M(hk ) there is h‘k ϵ M(ho ) such that h'k - hk → 0 as k → ∞ . If, in addition, the set M is finite or M is the convex hull of a finite set of elements, then the multivalued mapping h → M(h) is lower semicontinuous on coM. First published online: 14 Oct 2010
Convergence of the finite volume method for a conductive-radiative heat transfer problem
We show that the finite volume method rigorously converges to the solution of a conductive-radiative heat transfer problem with nonlocal and nonlinear boundary conditions. To get this result, we start by proving existence of solutions for a finite volume discretization of the original problem. Then, by obtaining uniform boundedness of discrete solutions and their discrete gradients with respect to mesh size, we finally get L 2type convergence of discrete solutions.
Relaxation of Quasilinear Elliptic SystemsviaA-quasiconvex Envelopes
We consider the weak closure WZof the set Z of all feasible pairs (solution, flow) of the family of potential elliptic systems div s0 s=1 s(x)F 0 s(ru(x )+ g(x)) f(x) =0i n; u =( u1;:::;um)2 H 1 0 (; R m ) ; =( 1;:::;s 0 )2 S; where R n is a bounded Lipschitz domain, Fs are strictly convex smooth functions with quadratic growth and S =f measurable j s(x )=0o r 1 ;s =1 ;:::;s0 ;1(x )+ +s0 (x )=1 g .W e show that WZis the zero level set for an integral functional with the integrand QF being the A-quasiconvex envelope for a certain functionF and the operator A = (curl,div) m . If the functions Fs are isotropic, then on the characteristic cone (dened by the operator A) QF coincides with the A-p…