Non-convex minimization -- the case of vector fields

From MaRDI portal





Making use of the theory of (weakly) monotone operators, the author studies minimization problems without global weak lower semicontinuity. The main result can be stated as follows. Let \(E\) be a real reflexive Banach space and \(f:E\to \mathbb{R}\) a continuous functional with semi-continuous and locally bounded Gateau derivative \(f':E\to E'\). If \(M\subset E\) is such that \((a)\) \(\inf_{M}f(u)>-\infty\), \((b)\) \(M\cap f^{-1}([a,b])\) is bounded in \(E\) for any \([a,b]\subset \mathbb{R}\), \((c)\) if \(u_{n}\to v\) weakly in \(E\), \((u_{n})\subset M\) and \(\lim f(u_{n})=\inf_{M}f(u)\), then \(v\in M\), \noindent then \(f\) admits a minimum on \(M\) if and only if there exist \((u_{n})\), \(v\in M\) such that \(u_{n}\to v\) weakly in \(E\), \(\lim f(u_{n})=\inf_{M}f(u)\) and (setting \(K_{\varepsilon}((u_{n}))=\{u\in E|u=t(u_{j}-v)\), \(\varepsilon\leq t\leq 1\), \(j\in\mathbb{N}\}\)) \(f'\) is weakly \(K_{\varepsilon}((u_{n}))\)-monotone at \(v\) whenever \(\varepsilon\in (0,\varepsilon_{0})\). The author then describes some `concrete' no quasi-convex situations which can be solved by the above stated abstract result announcing the proof in a forthcoming paper. For the entire collection see [Zbl 0970.00021].











This page was built for publication: Non-convex minimization -- the case of vector fields

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2760028)