ProbXiv
Explorersign in
Problem archiveProblem record

Statement

Assume the setup on pages 68–69: y′=M(⋅)yy'=M(\cdot)y has an exponential dichotomy on R\mathbb R with dichotomy projector QQ, and {Qj}j=1n\{Q_j\}_{j=1}^n is its ordered finest exponential splitting, with Q=Q1+⋯+QkQ=Q_1+\cdots+Q_k. Let R∈L1(R;Cd×d)R\in L^1(\mathbb R;\mathbb C^{d\times d}) and suppose that, for some δ∈(0,ε0)\delta\in(0,\varepsilon_0),

∫0∞e(κj+−κj′++δ)x∥R(x)∥ dx<∞(j=1,…,k),\int_0^\infty e^{(\kappa_j^+-\kappa_j^{\prime +}+\delta)x}\lVert R(x)\rVert\,dx<\infty\quad(j=1,\ldots,k), ∫−∞0e−(κj−−κj′−+δ)x∥R(x)∥ dx<∞(j=k+1,…,n).\int_{-\infty}^0 e^{-(\kappa_j^- -\kappa_j^{\prime -}+\delta)x}\lVert R(x)\rVert\,dx<\infty\quad(j=k+1,\ldots,n).

Let ε∈(0,δ)\varepsilon\in(0,\delta), and let {Q~j+}j=1n\{\widetilde Q_j^+\}_{j=1}^n and {Q~j−}j=1n\{\widetilde Q_j^-\}_{j=1}^n be exponential splittings for y′=[M(⋅)+R(⋅)]yy'=[M(\cdot)+R(\cdot)]y over R+\mathbb R_+ and R−\mathbb R_-, respectively, that are ε\varepsilon-close to {Qj}j=1n\{Q_j\}_{j=1}^n. Then, for every γ∈(ε,δ)\gamma\in(\varepsilon,\delta), the perturbed equation has absolutely continuous matrix solutions Yj(⋅)Y_j(\cdot), with j=1,…,kj=1,\ldots,k on R+\mathbb R_+ and j=k+1,…,nj=k+1,\ldots,n on R−\mathbb R_-, such that, for E~j+(x)=⨁i=1jIm⁡(Q~i+(x))\widetilde E_j^+(x)=\bigoplus_{i=1}^j\operatorname{Im}(\widetilde Q_i^+(x)) and E~j−(x)=⨁i=jnIm⁡(Q~i−(x))\widetilde E_j^-(x)=\bigoplus_{i=j}^n\operatorname{Im}(\widetilde Q_i^-(x)),

Yj(x)∈E~j+(x)∖E~j−1+(x),x≥0,Y_j(x)\in\widetilde E_j^+(x)\setminus\widetilde E_{j-1}^+(x),\quad x\ge0, sup⁡x≥0e−(κj′+−γ)x∥Yj(x)−Φ(x)Qj∥<∞,j=1,…,k,\sup_{x\ge0}e^{-(\kappa_j^{\prime +}-\gamma)x}\lVert Y_j(x)-\Phi(x)Q_j\rVert<\infty,\quad j=1,\ldots,k, Yj(x)∈E~j−(x)∖E~j+1−(x),x≤0,Y_j(x)\in\widetilde E_j^-(x)\setminus\widetilde E_{j+1}^-(x),\quad x\le0, sup⁡x≤0e−(κj−+γ)x∥Yj(x)−Φ(x)Qj∥<∞,j=k+1,…,n.\sup_{x\le0}e^{-(\kappa_j^-+\gamma)x}\lVert Y_j(x)-\Phi(x)Q_j\rVert<\infty,\quad j=k+1,\ldots,n.

These solutions are unique up to terms of lower exponential order. Moreover, if the splittings are those constructed in [DK, Theorem IV.5.1], there is a unique set of absolutely continuous matrix solutions satisfying Yj(⋅)∈Im⁡(Φ~(⋅)Q~j+Φ~(⋅)−1)Y_j(\cdot)\in\operatorname{Im}(\widetilde\Phi(\cdot)\widetilde Q_j^+\widetilde\Phi(\cdot)^{-1}) on R+\mathbb R_+ for j≤kj\le k, respectively Yj(⋅)∈Im⁡(Φ~(⋅)Q~j−Φ~(⋅)−1)Y_j(\cdot)\in\operatorname{Im}(\widetilde\Phi(\cdot)\widetilde Q_j^-\widetilde\Phi(\cdot)^{-1}) on R−\mathbb R_- for j>kj>k, and the same estimates hold for every γ∈(ε,δ)\gamma\in(\varepsilon,\delta).

Record

Comments

Nothing has been published against this problem yet, and nothing has been said about it.

Sign in with an institutional address to take part in the discussion. Reading every thread stays open to everyone.

Sign in

Solve with an agent

Open the statement in a chat, with the problem and the ground rules already written into the prompt.

This opens a third-party site. Nothing is posted back to ProbXiv and nothing you write there is recorded here — what a model gives you is an attempt, which a person still has to check.