Nonlinear second-order multivalued boundary value problems

Leszek Gasinski,Nikolaos S. Papageorgiou
DOI: https://doi.org/10.48550/arXiv.math/0310264
2003-10-17
Abstract:In this paper we study nonlinear second-order differential inclusions involving the ordinary vector $p$-Laplacian, a multivalued maximal monotone operator and nonlinear multivalued boundary conditions. Our framework is general and unifying and incorporates gradient systems, evolutionary variational inequalities and the classical boundary value problems, namely the Dirichlet, the Neumann and the periodic problems. Using notions and techniques from the nonlinear operator theory and from multivalued analysis, we obtain solutions for both the `convex' and `nonconvex' problems. Finally, we present the cases of special interest, which fit into our framework, illustrating the generality of our results.
Classical Analysis and ODEs,Analysis of PDEs
What problem does this paper attempt to address?
The problem that this paper attempts to solve is the research on nonlinear second - order multivalued boundary value problems. Specifically, the authors studied nonlinear second - order differential inclusions problems involving the ordinary vector p - Laplacian, multivalued maximal monotone operators, and nonlinear multivalued boundary conditions. Such problems can uniformly handle gradient systems, evolutionary variational inequalities, and classical boundary value problems (such as Dirichlet problems, Neumann problems, and periodic problems). The main goal of the paper is to find solutions to such problems under given assumptions, especially for "convex" and "non - convex" problems, by using the concepts and techniques of nonlinear operator theory and multivalued analysis. The specific form of the problem mentioned in the paper is as follows: \[ \begin{cases} \phi(x'(t))' \in A(x(t))+F(t, x(t)) & \text{almost everywhere in }[0, T]\\ (\phi(x'(0)), -\phi(x'(T))) \in \xi(x(0), x(T)) \end{cases} \] where: - $\phi:\mathbb{R}^N\rightarrow\mathbb{R}^N$ is defined as $\phi(\zeta)=\|\zeta\|^{p - 2}_{\mathbb{R}^N}\zeta$, $p\geq2$. - $A:\mathbb{R}^N\supseteq D(A)\rightarrow2^{\mathbb{R}^N}$ is a maximal monotone mapping. - $F:[0, T]\times\mathbb{R}^N\rightarrow2^{\mathbb{R}^N}$ is a multivalued vector field. - $\xi:\mathbb{R}^N\times\mathbb{R}^N\rightarrow2^{\mathbb{R}^N\times\mathbb{R}^N}$ describes the boundary conditions and is also a maximal monotone mapping. The authors derive a priori bounds of solutions by introducing the Hartman condition and finally apply the generalized multivalued Leray - Schauder alternative principle to prove the existence of solutions. The paper also discusses the applicability of this framework in special cases, showing the generality of the results.