Strong solutions for differential equations in abstract spaces (Q2484526): Difference between revisions

The author considers in a sequentially complete locally convex space \((E,{\mathcal F})\) the initial value problem \[ u_t(t)= f(t, u(t)),\quad u(t_0)= u_0. \] The bounded elements of \(E\) are denoted by \[ E_b:= \Biggl\{x\in E:\| x\|_{{\mathcal F}}= \sup_{\rho\in{\mathcal F}}\,\rho(x)< \infty\Biggr\}. \] \((E_b,{\mathcal F})\) is supposed to be locally metrizable, and \(B_{E_b}\) is supposed to be relatively compact with respect to the \({\mathcal F}\)-topology, \(f: I\times E_b\to E_b\) is an \({\mathcal F}\)-Carathéodory map (i.e., an extension of the classical concept) satisfying a further inequality. By using the Schauder-Tychonoff fixed-point theorem, the local existence of a strong solution (i.e., differentiable with respect to the \(\|\cdot\|_{{\mathcal F}}\)-norm) in \(E_b\) with \(x_0\in E_b\) is proved. The main tools used in this proof are developed in a preceding section. In the last section, the author studies by using the previously developed results a nonlinear differential equation involving the Hardy-Littlewood maximal operator and proves Lipschitz stability of the solution. I want to point out that the author makes an effort to explain his problem: In a long and interesting introduction, he gives a survey of the development of existence theorems for differential equations in Banach spaces. He compares assumptions and results and a long list of references is given. For his own problem, he gives motivations and discusses the assumptions.