\documentclass[reqno]{amsart} \usepackage{hyperref} \AtBeginDocument{{\noindent\small \emph{Electronic Journal of Differential Equations}, Vol. 2011 (2011), No. 117, pp. 1--12.\newline ISSN: 1072-6691. URL: http://ejde.math.txstate.edu or http://ejde.math.unt.edu \newline ftp ejde.math.txstate.edu} \thanks{\copyright 2011 Texas State University - San Marcos.} \vspace{9mm}} \begin{document} \title[\hfilneg EJDE-2011/117\hfil S-asymptotically periodic solutions] {S-asymptotically periodic solutions for partial differential equations with finite delay} \author[W. Dimbour, G. Mophou, G. M. N'Gu\'er\'ekata\hfil EJDE-2011/117\hfilneg] {William Dimbour, Gis\`ele Mophou, Gaston M. N'Gu\'er\'ekata} % in alphabetical order \address{William Dimbour\newline Laboratoire C.E.R.E.G.M.I.A., Universit\'e des Antilles et de la Guyane, Campus Fouillole 97159 Pointe-\`a-Pitre Guadeloupe (FWI)} \email{William.Dimbour@univ-ag.fr} \address{Gis\`ele Mophou \newline Laboratoire C.E.R.E.G.M.I.A., Universit\'e des Antilles et de la Guyane, Campus Fouillole 97159 Pointe-\`a-Pitre Guadeloupe (FWI)} \email{gisele.Mophou@univ-ag.fr} \address{Gaston M. N'Gu\'er\'ekata \newline Department of Mathematics, Morgan State University, 1700 East Cold Spring Lane, Baltimore, MD 21251, USA} \email{Gaston.N'Guerekata@morgan.edu, nguerekata@aol.com} \thanks{Submitted August 11, 2011. Published September 14, 2011.} \subjclass[2000]{34K05, 34A12, 34A40} \keywords{S-asymptotically periodic function; mild solution; \hfill\break\indent exponentially stable semigroup; fractional power operator} \begin{abstract} In this article, we give some sufficient conditions for the existence and uniqueness of S-asymptotically periodic (mild) solutions for some partial functional differential equations. To illustrate our main result, we study a diffusion equation with delay. \end{abstract} \maketitle \numberwithin{equation}{section} \newtheorem{theorem}{Theorem}[section] \newtheorem{lemma}[theorem]{Lemma} \newtheorem{corollary}[theorem]{Corollary} \newtheorem{proposition}[theorem]{Proposition} \newtheorem{remark}[theorem]{Remark} \newtheorem{definition}[theorem]{Definition} \allowdisplaybreaks \section{Introduction} The main purpose of this work is to study the existence and uniqueness of S-asymptotically periodic solutions in the $\alpha$-norm for the partial differential equation \begin{equation}\label{eq22} \begin{gathered} \frac{d}{dt} u(t)=-Au(t)+L(u_t)+f(t,u(t))\quad\text{for }t \ge 0,\\ u_0=\varphi \end{gathered} \end{equation} where $-A$ is the infinitesimal generator of an analytic semigroup $T(t)$, $t\geq 0$ on a Banach space $\mathbb{X}$. For $0<\alpha\le 1$, let $ A^{\alpha}$ be the fractional power of $A$ with domain $D(A^{\alpha})$, which endowed with the norm $|x|_{\alpha}=\|A^{\alpha}x\|$ forms a Banach space $\mathbb{X}_{\alpha}$. Let $\mathcal{C}_{\alpha}=C([-r,0],\mathbb{X}_\alpha)$ be the Banach space of all continuous functions from $[-r,0]$ to $\mathbb{X}_{\alpha}$ endowed with the norm $$ | \phi|_{\mathcal{C}_{\alpha}} =\sup_{-r\le\theta\le0}|\phi(\theta)|_{\alpha}. $$ Let $L$ be a bounded linear operator from $\mathcal{C}_{\alpha}$ to $\mathbb{X}_{\alpha}$, and $f:\mathbb{R}\times \mathbb{X}_\alpha\to \mathbb{X}_\alpha$ a continuous function. As usual the history function $x_t \in\mathcal{C}_{\alpha}$ is defined by $$ x_t(\theta)=x(t+\theta)\quad\text{for }\theta\in]-r,0]. $$ The theory of partial functional differential equations and its applications are an active are of research; see for instance \cite{Hale,traweb,Wu} and the references therein. Several articles study the existence and uniqueness of almost periodic, almost automorphic, and weighted pseudo almost periodic solutions of various differential equations. In \cite{Elazzouzi}, the author deals with the existence of $C^{(n)}$-almost periodic and $C^{(n)}$-automorphic solution of the equation \begin{equation}\label{eq20} \begin{gathered} \frac{d}{dt} u(t)=-Au(t)+L(u_t)+f(t)\quad\text{for }t \ge 0,\\ u_0=\varphi \end{gathered} \end{equation} To achieve his goal, the author uses the the variation of constants formula and the reduction method developed by Adimy et al. \cite{adimy0}. Ezzinbi and Boukli-Hacene \cite{Ezzinbi0} studied the existence and uniqueness of weighted pseudo-almost automorphic solution for \eqref{eq20}, using the variation of constants formula developed by Ezzinbi and N'Gu\'er\'ekata \cite{Ezzinbi}. The literature relative to S-asymptotically periodic functions remains limited due to the novelty of the concept. Qualitative properties of such functions are discussed for instance in \cite{Blot,Henriquez1,Lizama}. In \cite{Blot}, the authors present a new composition theorem for such functions. Various properties of S-asymptotically periodic functions are also investigated in a general study of classes of bounded continuous functions taking values in a Banach space $\mathcal{X}$. In \cite{Cuevas}, a new concept of weighted S-asymptotically periodic functions is introduced generalizing in a natural way the one studied here. There are some papers dealing with the existence of S-asymptotically periodic solutions of differential equations and fractional differential equations in finite as well as infinite dimensional spaces; see \cite{Blot,Henriquez1,Henriquez2,Lizama,Nicola}. In this paper, motivated by all these works, we first reconsider \eqref{eq20} and prove that if $f$ is an S-asymptotically periodic function in the $\alpha$-norm then its has a unique solution on $[-r,+\infty[$ . Moreover, the restriction of the solution on $\mathbb{R}^+$ is S-asymptotically periodic solutions in the $\alpha$-norm. This allow us to study the existence and uniqueness of an S-asymptotically periodic solution in the $\alpha$-norm, for \eqref{eq22}. This work is organized as follows. In Section \ref{prem}, we recall some fundamental properties of S-asymptotically periodic functions and fractional powers of a closed operator. Section \ref{main} is devoted to the main result. We illustrate our main result in Section \ref{appl} by examining the existence and uniqueness of S-asymptotically periodic (mild) solutions for some diffusion equations with delay. \section{PRELIMINARIES \label{prem}} Let $(\mathbb{X},\|\cdot\|)$ be a Banach space. Denote by $C(\mathbb{R}^+, \mathbb{X})$, the space of all continuous functions from $\mathbb{R}^+$ to $ \mathbb{X}$, and by $BC(\mathbb{R}^+ , \mathbb{X})$ the space of all bounded continuous functions $\mathbb{R}^+ \to \mathbb{X}$. The space $BC(\mathbb{R}^+ , \mathbb{X})$ endowed with the supremum norm $\| f \|_{\infty}:=\sup_{t \ge 0} | | f(t) | |$ is a Banach space. \subsection*{S-asymptotically periodic functions} \begin{definition} \rm For a function $f$ in $BC(\mathbb{R}^+, \mathbb{X})$, we say that $f$ belongs to $C_0(\mathbb{R}^+, \mathbb{X})$ if $ \lim_{ t\to \infty}\| f(t)\|=0$. Let $\omega$ be a fixed positive number and $f\in BC(\mathbb{R}^+, \mathbb{X})$. We say that $f$ is $\omega$-periodic, denoted by $f \in P_{\omega}(\mathbb{X})$, if $f$ has period $\omega$. Note that $P_{\omega}(\mathbb{X})$ is a Banach subspace of $BC(\mathbb{R}^+ , \mathbb{X})$ under the supremum norm. \end{definition} \begin{definition}[\cite{Blot,Lizama}] \rm Let $f \in BC(\mathbb{R}^+ , \mathbb{X})$ and $\omega>0$. We say that $f$ is asymptotically $\omega$-periodic if $f=g+h$ where $g \in P_{\omega}(\mathbb{X})$ and $h\in C_0(\mathbb{R}^+ ,\mathbb{X})$. We denote by $AP_{\omega}(\mathbb{X})$ the set of all asymptotically $\omega$-periodic functions from $\mathbb{R}^+$ to $\mathbb{X}$. Note that $AP_{\omega}(\mathbb{X})$ is a Banach space under the supremum norm. \end{definition} From the above definitions, it follows that $AP_{\omega}(\mathbb{X})=P_{\omega}(\mathbb{X}) \bigoplus C_0(\mathbb{R}^+ ,\mathbb{X})$; cf. \cite{Lizama}. \begin{definition}[\cite{Henriquez1}] \rm A function $f \in BC(\mathbb{R}^+ , \mathbb{X})$ is called S-asymptotically $\omega$-periodic if there exists $\omega$ such that $\lim_{ t \to \infty}(f(t+\omega)-f(t))=0$. In this case we say that $\omega$ is an asymptotic period of $f$ and that $f$ is S-asymptotically $\omega$-periodic. \end{definition} We will denote by $SAP_{\omega}(\mathbb{X})$, the set of all S-asymptotically $\omega$-periodic functions from $\mathbb{R}^+ to \mathbb{X}$. Then we have $$ AP_{\omega}(\mathbb{X}) \subset SAP_{\omega}(\mathbb{X}). $$ Note that the inclusion above is strict. Consider the function $f:\mathbb{R}^+ \to c_0$ where $c_0=\{ x=(x_n)_{n \in \mathbb{N}}: \lim_{n \to \infty}x_n=0 \}$ equipped with the norm $\| x \| = \sup_{n \in \mathbb{N}} | x(n) |$, and $f(t)= (\frac{2nt^2}{t^2 + n^2})_{n \in \mathbb{N}}$. Then $f \in SAP_{\omega}(c_0)$ but $ f \notin AP_{\omega}(c_0)$; see \cite[Example 3.1]{Henriquez1}. The following result is due to Henriquez-Pierri-T\`aboas; \cite[Proposition 3.5]{Henriquez1}. \begin{theorem} \label{thm1} The space $SAP_{\omega}(\mathbb{X})$ endowed with the norm $\| \cdot \|_{\infty}$ is a Banach space. \end{theorem} \begin{theorem}[{\cite[Theorem 3.7]{Blot}}] \label{thm2} Let $\phi:\mathbb{X} \to \mathbb{Y}$ be a function which is uniformly continuous on bounded subsets of $\mathbb{X}$ and such that $\phi$ maps bounded subsets of $\mathbb{X}$ into bounded subsets of $\mathbb{Y}$. Then for all $f\in SAP_{\omega}(\mathbb{X})$, the composition $\phi \circ f:=[t\to\phi(f(t))]\in SAP_{\omega}(\mathbb{X})$. \end{theorem} \begin{corollary}[{\cite[Corollary 3.10]{Blot}}] \label{coro} Let $\mathbb{X}$ and $\mathbb{Y}$ be two Banach spaces, and denote by $ \mathbb{B}(\mathbb{X},\mathbb{Y})$, the space of all bounded linear operators from $\mathbb{X}$ into $\mathbb{Y}$. Let $A \in \mathbb{B}(\mathbb{X},\mathbb{Y})$. Then when $f \in SAP_{\omega}(\mathbb{X})$, we have $Af:=[t \to Af(t)] \in SAP_{\omega}(\mathbb{Y})$. \end{corollary} Next we consider asymptotically $\omega$-periodic functions with parameters. \begin{definition}[\cite{Henriquez1}] \label{def5} A continuous function $f:[0,\infty[\times \mathbb{X} \to \mathbb{X}$ is said to be uniformly S-asymptotically $\omega$-periodic on bounded sets if for every bounded set $K \subset \mathbb{X}$, the set $\{f(t,x):t\ge 0,x \in K \}$ is bounded and $ \lim_{ t \to \infty}(f(t,x)-f(t+\omega, x))=0$ uniformly in $x \in K$. \end{definition} \begin{definition}[\cite{Henriquez1}] \label{def6} \rm A continuous function $f:[0,\infty[\times \mathbb{X} \to \mathbb{X}$ is said to be asymptotically uniformly continuous on bounded sets if for every $\epsilon>0$ and every bounded set $K \subset \mathbb{X}$, there exist $L_{\epsilon, K}>0$ and $\delta_{\epsilon, K}>0$ such that $| | f(t,x)-f(t,y)\|<\epsilon$ for all $t\ge L_{\epsilon, K}$ and all $x,y \in K$ with $\| x-y \| < \delta_{\epsilon, K}$. \end{definition} \begin{theorem}[\cite{Henriquez1}] \label{thm3} Let $f:[0,\infty[\times \mathbb{X} \to \mathbb{X}$ be a function which uniformly S-asymptotically $\omega$-periodic on bounded sets and asymptotically uniformly continuous on bounded sets. Let $u:[0,\infty[$ be S-asymptotically $\omega$-periodic function. Then the Nemytskii operator $\phi(\cdot):=f(\cdot,u(\cdot))$ is S-asymptotically $\omega$-periodic function. \end{theorem} \subsection*{Fractional powers of the operator $A$} Let $\varrho (A)$ denote the resolvent set of $A$. We assume without loss of generality that \begin{equation}\label{rho0} 0\in \varrho(A). \end{equation} This allows us, on the one hand, to say that there exist constants $M>1$ and $\delta>0$ such that \begin{equation}\label{bornT} \| T(t) x\|\le Me^{-\delta t}\| x\| ,\quad \forall t \ge 0,\; x \in\mathbb{X}\,, \end{equation} and on the other hand, to define the fractional power $A^\alpha$ for $0 <\alpha <1$, as a closed linear operator on its domain $ D(A^\alpha)$ with inverse $A^{-\alpha} $ given by $$ A^{-\alpha}=\frac{1}{\Gamma(\alpha)}\int_0^t t^{\alpha-1}T(t)dt $$ where $\Gamma$ denotes the Gamma function $$ \Gamma(\alpha)=\int_0^t t^{\alpha-1}e^{-\alpha t}dt. $$ We have the following basic properties for $A^{\alpha}$. \begin{theorem}[{\cite[pp. 69-75]{Pazy}}] \label{thm4} For $0<\alpha< 1$, the following properties hold. \begin{itemize} \item[(i)] $\mathbb{X}_\alpha=D(A^{\alpha})$ is a Banach space with the norm $|x|_\alpha=\| A^\alpha x\| $ for $ x\in D(A^{\alpha})$; \item[(ii)] $A^{-\alpha}$ is the closed linear operator with $Im(A^{-\alpha})=D(A^{\alpha})$ and we have $A^{\alpha}=(A^{-\alpha})^{-1}$; \item[(iii)] $A^{-\alpha}\in \mathbb{B}(\mathbb{X},\mathbb{X})$; \item[(iv)] $T(t):\mathbb{X}\to \mathbb{X}_{\alpha}$ for every $t>0$; \item[(v)] $A^{\alpha}T(t)x=T(t)A^{\alpha}x$ for each $x \in D(A^{\alpha})$ and $t\ge0$; \item[(vi)] $0<\alpha\le\beta$ implies $D(A^{\beta})\hookrightarrow D(A^{\alpha})$; \item[(vii)] There exists $M_\alpha>1$ such that $$ \| A^{\alpha}T(t)x| | \le M_{\alpha}\frac{e^{-\delta t}}{t^{\alpha}} \| x\|\quad\text{for }x \in \mathbb{X},\; t>0. $$ where $\delta >0$ is given by \eqref{bornT} \end{itemize} \end{theorem} \begin{remark} \label{rmk1}\rm Observe as in \cite{hsiang,Mophou} that from Theorem \ref{thm4} (iv) and (v), the restriction $T_\alpha(t)$ of $T(t)$ to $\mathbb{X}_\alpha$ is exactly the part of $T(t)$ in $\mathbb{X}_\alpha$. Let $x \in\mathbb{X}_{\alpha}$. $$ | T(t)x|_{\alpha}=\| A^{\alpha}T(t)x\| =\| T(t)A^{\alpha} x\|\le | T(t)|\,| | A^{\alpha} x\| =| T(t)|\, | x|_{\alpha}, $$ and as $t$ decreases to $0$, $$ | T(t)x - x|_\alpha=\| A^\alpha T(t)x - A^\alpha x\| = \| T(t)A^\alpha x - A^\alpha x\| \to 0, $$ for all $x\in \mathbb{X}_\alpha;$ it follows that $(T(t))_{t\geq 0}$ is a family of strongly continuous semigroup on $\mathbb{X}_\alpha$ and $| T_\alpha (t)| \leq | T(t)|$ for all $t\geq 0$. \end{remark} \begin{proposition}[\cite{Elazzouzi,Travis}] $((T(t)_{t\geq 0})$ is a strongly continuous semigroup on $\mathcal{C}_\alpha$; that is, \begin{itemize} \item[(i)] for all $t\geq 0$ $T(t)$ is a bounded linear operator on $\mathcal{C}_\alpha$; \item[(ii)] $T(0)=I$; \item[(iii)] $T(t+s)=T(t)T(s)$ for all $t,\, s \geq 0$; \item[(iv)] for all $\varphi \in \mathcal{C}_\alpha$, $T(t) \varphi$ is a continuous function of $t\geq 0$ with values in $\mathcal{C}_\alpha$. \end{itemize} \end{proposition} \section{Applications to partial differential equations with finite delay} \label{main} \begin{definition}\label{def7} \rm Let $ \varphi\in \mathcal{C}_{\alpha}$. A function $u:[-r, +\infty[\to\mathbb{X}_{\alpha}$ is said to be a mild solution of \eqref{eq20} if the following conditions hold: \begin{itemize} \item[(i)] $u:[-r, +\infty[\to\mathbb{X}_{\alpha}$ is continuous; \item[(ii)] $u(t)= T(t)\varphi(0)+\int_{0}^t T(t-s)[L(u_s)+f(s)]ds$ for $t\ge 0$; \item[(iii)] $u_{0}=\varphi$. \end{itemize} \end{definition} For the rest of this article, we define $$ \Omega=\{u:[-r,\,+\infty[\to \mathbb{X}_\alpha\text{ such that } u|_{[-r,0]}\in \mathcal{C}_\alpha \text{ and } u|_{\mathbb{R}^+}\in SAP_{\omega}(\mathbb{X}_{\alpha}) \}. $$ Note that if $u\in \Omega$ then $u $ is bounded on $[-r,\,+\infty[$. We set \begin{equation}\label{bornu} \| u\|_\Omega = {\sup_{ s\in [-r,+\infty[}}| u(s)|_\alpha. \end{equation} It is clear that $\| u\|_\infty \leq \| u\|_\Omega $. \begin{lemma}\label{lem1} Under assumption \eqref{rho0}, the function $l$ defined by $$ l(t)=T(t)\varphi(0) $$ belongs to $SAP_{\omega}(\mathbb{X}_{\alpha})$. \end{lemma} \begin{proof} Since $\varphi(0)\in \mathbb{X}_\alpha$, we have on the one hand that $(T(t))_{t\geq 0}$ is a family of strongly continuous semigroup on $\mathbb{X}_\alpha$ (see Remark \ref{rmk1}), and on the other hand that $| l(t)|_\alpha\leq M|\varphi(0)|_\alpha$ because \eqref{bornT} holds. Consequently $l\in BC(\mathbb{R}^+,\mathbb{X}_\alpha)$. Now using \eqref{bornT} and Remark \ref{rmk1}, we obtain for $t\geq 0$, \begin{align*} | l(t+\omega)-l(t)|_{\alpha} &= |T(t+\omega)\varphi(0)-T(t)\varphi(0)|_{\alpha}\\ &\le |T(t+\omega)\varphi(0)|_{\alpha}+ | T(t)\varphi(0)|_{\alpha}\\ &\le | T(t+\omega)| | \varphi(0)|_{\alpha}+| T(t)| | \varphi(0)|_{\alpha}\\ &\le Me^{-\delta (t+\omega)}| \varphi(0)|_{\alpha}+Me^{-\delta t}| \varphi(0)|_{\alpha}. \end{align*} As $\delta >0$, we deduce that $$ \lim_{t \to \infty}| l(t+\omega)-l(t)|_{\alpha} = 0. $$ Thus $l \in SAP_{\omega}(\mathbb{X}_{\alpha})$. \end{proof} \begin{lemma}\label{lem2} If $u \in \Omega$, then \begin{gather} | u_t|_{\mathcal{C}_\alpha} \leq \| u\|_\Omega,\label{est1}\\ | L(u_t)|_\alpha \leq | L|_{\mathbb{B}(\mathcal{C}_\alpha, \mathbb{X}_\alpha)}\| u\|_\Omega\label{est2}\\ {\lim_{ t\to +\infty}}| u_{t+\omega}-u_t|_{\mathcal{C}_\alpha} = 0 \label{est3}. \end{gather} \end{lemma} \begin{proof} For any $\theta\in [-r,0]$ and $t\geq 0$, we have $$ | u_{t}(\theta)|_{\alpha}=| u(t+\theta)|_{\alpha}. $$ Since $u_t$ is continuous on $[-r,0]$ which is compact, we know that there exists $ \theta^* \in [-r,0]$ such that \[ | u_{t}|_{\mathcal{C}_\alpha} = \sup_{ -r\le\theta\le0}| u(t+\theta)|_{\alpha} = | u(t+\theta^*)|_{\alpha}. \] Since $u \in \Omega$, we deduce that \eqref{est1} holds. As $L\in \mathbb{B}(\mathcal{C}_\alpha,\mathbb{X}_\alpha)$, we can write $$ | L( u_{t})|_{\mathbb{X}_\alpha} \leq | L|_{\mathbb{B}(\mathcal{C}_\alpha,\mathbb{X}_\alpha)} | u_t|_{\mathcal{C}_\alpha}. $$ Therefore, using \eqref{est1}, we obtain \eqref{est2}. To complete the proof of the lemma, it suffices to prove \eqref{est3}. As $u_t$ is continuous on $[-r,0]$ which is compact, there exists $ \theta^* \in [-r,0]$ such that \begin{align*} | u_{t+\omega}-u_{t}|_{\mathcal{C}_\alpha} &= {\sup_{ -r\le\theta\le0}}| u(t+\theta+\omega)-u(t+\theta)|_{\alpha}\\ &= | u(t+\theta^*+\omega)-u(t+\theta^*)|_{\alpha}. \end{align*} Set $s=t+\theta$. Then, as $t$ tends to $+\infty$ we have $s$ tends to $+\infty$. Consequently $$ {\lim_{ t \to \infty}} | u(t+\theta^*+\omega)-u(t+\theta^*)|_{\alpha} = {\lim_{ s \to \infty}}| u(s+\omega)-u(s)|_{\alpha} = 0 $$ since $u \in \Omega$. Hence, $\lim_{t \to \infty}| u_{t+\omega}-u_{t}|_{\mathcal{C}_\alpha}=0$. \end{proof} \begin{lemma}\label{lem3} Assume that \eqref{rho0} holds. Let $f \in SAP_{\omega}(\mathbb{X}_{\alpha})$ and $\phi \in \Omega $. Then the function $\Phi: t \mapsto L(\phi_t)+f(t)$ belongs to $SAP_{\omega}(\mathbb{X}_{\alpha})$. \end{lemma} \begin{proof} It is clear that $\Phi\in C(\mathbb{R}^+, \mathbb{X}_\alpha)$. Using Lemma \ref{lem2}, we obtain \[ | \Phi(t)|_\alpha \leq | L(\phi_t)|_\alpha+| f(t)|_\alpha \leq | L|_{\mathbb{B}(\mathcal{C}_\alpha,\mathbb{X}_\alpha)} \| \phi\|_\Omega +\| f\|_{\infty}. \] This implies that $\Phi\in BC(\mathbb{R}^+, \mathbb{X}_\alpha)$. Hence \begin{equation}\label{est4} \| \Phi\|_\infty\leq | L|_{\mathbb{B}(\mathcal{C}_\alpha, \mathbb{X}_\alpha)} | | \phi\|_\Omega+\| f\|_{\infty}. \end{equation} On the other hand, for all $t\geq 0$, \begin{align*} | \Phi_{t+\omega}-\Phi_{t}|_{\alpha} &\leq | L( \phi_{t+\omega}-\phi_{t})|_{\alpha}+| f(t+\omega) -f(t)|_{\alpha}\\ &\leq | L|_{\mathbb{B}(\mathcal{C}_\alpha,\mathbb{X}_\alpha)}| \phi_{t+\omega}-\phi_{t}|_{\mathcal{C}_\alpha}+| f(t+\omega) -f(t)|_{\alpha}, \end{align*} Since $\phi\in \Omega$, using Lemma \ref{lem2}-\eqref{est3} and the fact that $f\in SAP_{\omega}(\mathbb{X}_{\alpha})$, we deduce that \begin{equation}\label{lim1} {\lim_{ t \to \infty}}| \Phi_{t+\omega}-\Phi_{t}|_{\alpha}=0. \end{equation} This completes the proof. \end{proof} \begin{proposition}\label{prop2} Assume that \eqref{rho0} holds. Let $f \in SAP_{\omega}(\mathbb{X}_{\alpha})$. For each $\phi \in \Omega $, define the nonlinear operator $\wedge_0$ by $$ (\wedge_0 \phi)(t)=\begin{cases} \varphi(t) & \text{if } t\in [-r,0],\\ T(t)\varphi(0)+ \int_{0}^t T(t-s)[L(\phi_s)+f(s)]ds &\text{if } t\geq 0. \end{cases} $$ Then $\wedge_0$ maps $\Omega$ into itself. \end{proposition} \begin{proof} It is clear that $(\wedge_0 \phi)$ is defined on $[-r,+\infty[$ and because $\varphi \in \mathcal{C}_\alpha$, we have $(\wedge_0 \phi)|_{[-r,0]}\in \mathcal{C}_\alpha$. Thus it suffices to show that the function $$ v: t \to \int_{0}^t T(t-s)[L(\phi_s)+f(s)]ds \in SAP_{\omega}(\mathbb{X}_{\alpha})$$ to complete the proof, since by Lemma \ref{lem1}, $T(t)\varphi(0)\in SAP_{\omega}(\mathbb{X}_{\alpha})$. For $t\ge0$, let $\Phi(t)=L(\phi_t)+f(t)$. Then \begin{align*} v(t+\omega)-v(t) &= \int_0^{t+\omega}T(t+\omega-s)\Phi(s)\,ds-\int_0^tT(t-s)\Phi(s)\,ds\\ &= \int_0^{\omega}T(t+\omega-s)\Phi(s)\,ds +\int_{\omega}^{t+\omega}T(t+\omega-s)\Phi(s)\,ds\\ &\quad- \int_0^tT(t-s)\Phi(s)\,ds. \end{align*} Then $$ | v(t+\omega)-v(t)|_{\alpha} \le | I_1(t)|_{\alpha} + | I_2(t)|_{\alpha}, $$ where \begin{gather*} I_1(t)= \int_0^{\omega}T(t+\omega-s)\Phi(s)\,ds,\\ I_2(t)= \int_{\omega}^{t+\omega}T(t+\omega-s)\Phi(s)\,ds -\int_0^tT(t-s)\Phi(s)\,ds,\\ |I_1(t)|_{\alpha}= \big| \int_0^{\omega}T(t+\omega-s)\Phi(s)\,ds \big|_{\alpha} \leq \int_0^{\omega}| T(t+\omega-s)\Phi(s)|_{\alpha}ds \end{gather*} Since \begin{align*} \int_0^{\omega}| T(t+\omega-s)\Phi(s)|_{\alpha}ds &= \int_0^{\omega}| | A^\alpha T(t+\omega-s)\Phi(s)| | ds\\ &= \int_0^{\omega}| | T(t+\omega-s) A^\alpha\Phi(s)| | ds\\ &\leq \int_0^{\omega}Me^{-\delta (t+\omega-s)}\| A^{\alpha} \Phi(s)\| \,ds, \end{align*} using \eqref{est4} we deduce that \begin{align*} | I_1(t)|_{\alpha} &\leq Me^{-\delta (t+\omega)} \int_0^{\omega}e^{\delta s} | \Phi(s)| _\alpha ds\\ &\leq Me^{-\delta (t+\omega)}\| \Phi\|_\infty \int_0^{\omega}e^{\delta s} ds\\ &\leq {\frac{1}{\delta}}Me^{-\delta (t+\omega)}\| \Phi\|_\infty (e^{\delta w} -1)\\ &\leq {\frac{1}{\delta}}M\| \Phi\|_\infty e^{-\delta t} \end{align*} Consequently, $ \lim_{ t \to \infty}| I_1(t)|_{\alpha}=0$ In view of \eqref{lim1}, we can find $T_{\epsilon}$ sufficiently large such that $$ | \Phi(t+\omega)-\Phi(t)|_{\alpha} < \frac{\delta}{M}\epsilon, \quad\text{for } t>T_{\epsilon}. $$ After a change of variable, we obtain $$ I_2(t)= \int_0^t T(t-s) (\Phi(s + \omega)-\Phi(s))\,ds . $$ Thus we obtain $$ | I_2(t) |_{\alpha} \le \big| \int_0^{T_{\epsilon}}T(t-s) (\Phi(s + \omega)-\Phi(s))\,ds\big|_{\alpha} +\big| \int_{T_{\epsilon}}^{t}T(t-s) (\Phi(s + \omega)-\Phi(s))\,ds \big|_{\alpha}. $$ Observing that \begin{align*} \big| \int_0^{T_{\epsilon}}T(t-s) (\Phi(s + \omega) -\Phi(s))\,ds\big|_{\alpha} &\le \int_0^{T_{\epsilon}}\big|T(t-s)\big(\Phi(s + \omega) -\Phi(s) \big)\big|_{\alpha}ds\\ &\leq \int_0^{T_{\epsilon}} M e^{-\delta (t-s)} | \Phi(s + \omega)-\Phi(s) |_\alpha ds\\ &\leq 2 \int_0^{T_{\epsilon}} M e^{-\delta (t-s)} \| \Phi \|_{\infty} ds\\ &\leq 2M\|\Phi| |_{\infty} e^{-\delta t} \int_0^{T_{\epsilon}}e^{\delta s}ds\\ &\leq 2M \| \Phi| |_{\infty} e^{-\delta t} \big(\frac{e^{\delta T_{\epsilon} }}{\delta}-\frac{1}{\delta}\big), \end{align*} we deduce that $$ \lim_{t \to \infty}\big| \int_0^{T_{\epsilon}}T(t-s) (\Phi(s + \omega)-\Phi(s))\,ds\big|_{\alpha} =0 $$ since $ \lim_{t \to \infty} [ 2M \| \Phi| |_{\infty} e^{-\delta t} (\frac{e^{\delta T_{\epsilon} }}{\delta}-\frac{1}{\delta})] =0$. Also we have \begin{align*} \big| \int_{T_{\epsilon}} ^{t}T(t-s) (\Phi(s + \omega)-\Phi(s))\,ds \big|_{\alpha} & \le \int_{T_{\epsilon}} ^{t} \big| T(t-s) (\Phi(s + \omega)-\Phi(s))\big|_{\alpha}ds\\ &\le \int_{T_{\epsilon}} ^{t}| T(t-s)| |(\Phi(s + \omega)-\Phi(s))|_{\alpha} ds\\ &\le \int_{T_{\epsilon}}^t Me^{-\delta(t-s)} \frac{\delta}{M}\epsilon \leq \epsilon. \end{align*} Therefore $$ {\lim_{t \to \infty}} \int_{T_{\epsilon}} ^{t}T(t-s) (\Phi(s + \omega)-\Phi(s))\,ds =0. $$ Finally, we obtain ${\lim_{t \to \infty}} I_2(t) =0$ and we have $t \to \int_{0}^t T(t-s)[L(\phi_s)+f(s)]ds \in SAP_{\omega}(\mathbb{X}_{\alpha})$. In summary, we have proved that \begin{itemize} \item $(\wedge_0 \phi)$ is defined $[-r,+\infty[$, \item $(\wedge_0 \phi)|_[-r,0]\in \mathcal{C}_\alpha$, \item $(\wedge_0 \phi)|_{\mathbb{R}^+}\in SAP_{\omega} (\mathbb{X}_{\alpha})$; \end{itemize} that is, $(\wedge_0 \phi)\in \Omega$. \end{proof} \begin{theorem} \label{thm5} Suppose that \eqref{rho0} holds and $f \in SAP_{\omega}(\mathbb{X}_{\alpha})$. Let $v$ be the restriction of the mild solution of \eqref{eq20} on $\mathbb{R}^+$. Then $v \in SAP_{\omega}(\mathbb{X}_{\alpha})$. \end{theorem} \begin{proof} According to the definition of mild solution of \eqref{eq20} given by Definition \ref{def7}, we have for any $t\geq 0$, $$ v(t)=T(t)\varphi(0)+\int_{0}^t T(t-s)[L(u_s)+f(s)]ds. $$ Hence it suffices to apply Proposition \eqref{prop2}, with $u=\phi$, to obtain that $v$ belongs to $SAP_{\omega}(\mathbb{X}_{\alpha})$. \end{proof} We make the following assumption. \begin{itemize} \item[(H1)] The function $g:{R}^+ \times \mathbb{X}_{\alpha} \to \mathbb{X}_{\alpha}$, $t \to g(t,u)$ is continuous and there exists a constant $K_f \ge 0$ such that $$ | g(t,u)-g(t,v)|_\alpha \le K_g | u-v |_\alpha \quad \text{for all }t \in \mathbb{R}^+ \; ( u,v )\in \mathbb{X}^2. $$ \item[(H2)] $M\big(| L|_{\mathbb{B}(\mathcal{C}_\alpha, \mathbb{X}_\alpha)}+K_g\big)/\delta<1$. \end{itemize} \begin{definition}\label{def8} Let $ \varphi\in\mathcal{C}_{\alpha}$. A function $u:[-r, +\infty[\to\mathbb{X}_{\alpha}$ is said to be a mild solution of \eqref{eq22} if the following conditions hold: \begin{itemize} \item[(i)] $u:[-r, +\infty[\to\mathbb{X}_{\alpha}$ is continuous; \item[(ii)] $u(t)= T(t)\varphi(0)+ \int_{0}^t T(t-s)[L(u_s) +g(s,u(s))]ds \quad\text{for }t\ge 0$; \item[(iii)] $u_{0}=\varphi$. \end{itemize} \end{definition} \begin{proposition}\label{prop1} Suppose that \eqref{rho0} holds. Assume also that the function $g$ is uniformly S-asymptotically $\omega$-periodic on bounded sets and {\rm(H1)} hold. For each $\phi \in \Omega$, define the nonlinear operator $\wedge_1$ by $$ (\wedge_1 \phi)(t)= \begin{cases} \varphi(t) & \text{if } t\in [-r,0],\\ T(t)\varphi(0)+ \int_{0}^t T(t-s)[L(\phi_s)+g(s,\phi(s))]ds &\text{if } t\geq 0. \end{cases} $$ Then $\wedge_1$ maps $\Omega$ into itself. \end{proposition} \begin{proof} We have $\phi|_{\mathbb{R}^+} \in SAP_{\omega}(\mathbb{X}_{\alpha})$ since $\phi\in \Omega$. Since $g$ satisfying (H1), it follows from Theorem \ref{thm3} that the function $h: t\mapsto g(t,\phi(t))$ belongs to $ SAP_{\omega}(\mathbb{X}_{\alpha})$. Hence, it suffices to proceed exactly as for the proof of the Proposition \ref{prop2} replacing $f(\cdot)$ by $h(\cdot)$ to obtain that $\wedge_1$ maps $\Omega$ into itself. \end{proof} \begin{theorem} \label{thm6} Suppose that \eqref{rho0} and {\rm(H2)} hold. Also assume that the function $g$ is uniformly S-asymptotically $\omega$-periodic on bounded sets and {\rm (H1)} hold. Then for all $\varphi \in \mathcal{C}_{\alpha}$, Equation \eqref{eq22} has a unique mild solution in $\Omega$. \end{theorem} \begin{proof} Consider the operator $Q: \Omega \to \Omega$ defined by: $$ (Q u)(t)= \begin{cases} \varphi(t) & \text{if } t\in [-r,0],\\ T(t)\varphi(0)+ \int_{0}^t T(t-s)[L(u_s)+g(s,u(s))]ds &\text{if } t\geq 0. \end{cases} $$ Observe that in view of Proposition \ref{prop1}, $Q$ is well defined. Consider $u,v\in \Omega$. For all $t\in [-r,\,+\infty[$, we have \begin{align*} &| (Qu)(t)-(Qv)(t)|_\alpha\\ &=\big| \int_{0}^t T(t-s)[(L(u_s)-L(v_s))+(g(s,u(s))-g(s,v(s)))]ds \big|_\alpha\\ &\leq \int_{0}^t \big| T(t-s)[(L(u_s)-L(v_s))+(g(s,u(s))-g(s,v(s)))] \big|_\alpha ds. \end{align*} Therefore, using \eqref{bornT} and \eqref{est1}, we obtain \begin{align*} &| (Qu)(t)-(Qv)(t)|_\alpha\\ &\leq \int_{0}^t Me^{-\delta(t-s)} [| L(u_s)-L(v_s)|_\alpha +| g(s,u(s))-g(s,v(s))|_\alpha] ds \\ &\leq Me^{-\delta t}| L|_{\mathbb{B}(\mathcal{C}_\alpha, \mathbb{X}_\alpha)} \int_{0}^t e^{\delta s}| u_s - v_s|_{\mathcal{C}_\alpha} ds\\ &\quad + Me^{-\delta t}K_g \int_{0}^t e^{\delta s}| u(s) -v(s)|_\alpha ds \\ &\leq Me^{-\delta t}| L|_{\mathbb{B}(\mathcal{C}_\alpha, \mathbb{X}_\alpha)}\| u-v\|_{\Omega} \int_{0}^t e^{\delta s} ds\\ &\quad + Me^{-\delta t}K_g\| u-v\|_\infty \int_{0}^t e^{\delta s} ds. \end{align*} Since $\| u-v\|_\infty\leq \| u-v\|_{\Omega}$, we deduce that for all $t\geq -r$, \begin{align*} | (Qu)(t)-(Qv)(t)|_\alpha &\leq Me^{-\delta t}| L|_{\mathbb{B}(\mathcal{C}_\alpha,\mathbb{X}_\alpha)}\| u-v\|_{\Omega} \int_{0}^t e^{\delta s} ds\\ &\quad + Me^{-\delta t}K_g\| u-v\|_\infty \int_{0}^t e^{\delta s} ds\\ &\leq {\frac{Me^{-\delta t}}{\delta}} \big(| L|_{\mathbb{B}(\mathcal{C}_\alpha,\mathbb{X}_\alpha)}+K_g\big) \| u-v\|_{\Omega}(e^{\delta t}-1)\\ &\leq {\frac{M}{\delta}}\big(| L|_{\mathbb{B} (\mathcal{C}_\alpha,\mathbb{X}_\alpha)}+K_g\big)\| u-v\|_{\Omega}. \end{align*} Hence $$ \| (Qu)(t)-(Qv)(t)\|_\Omega\leq {\frac{M}{\delta}} \big(| L|_{\mathbb{B}(\mathcal{C}_\alpha,\mathbb{X}_\alpha)} +K_g\big)\| u-v\|_{\Omega}. $$ Hence assumption (H2) allows us to conclude in view of the contraction mapping principle that Q has a unique point fixed in $u\in \Omega$. The proof is now complete. \end{proof} \section{Application \label{appl}} Consider the functional partial differential equation \begin{equation}\label{syst2} \begin{gathered} \frac{\partial }{\partial t}u(t,x) =\frac{\partial ^2}{\partial x^2}u(t,x) +\int_{-r}^0 q(\theta)y(t+\theta,x)d\theta+g(t ,u(t,x))\quad t\in \mathbb{R}^+, x\in [0,\pi]\\ u(t,0)=u(t,\pi)=0\quad t\in\mathbb{R}^+\\ u(\theta,x)=\phi(\theta,x),\quad\text{for $\theta\in[-r,0]$ and $x\in[0,\pi]$} \end{gathered} \end{equation} where $q:[-r,0]\to\mathbb{R}$ is continuous. To study this system in the abstractt form \eqref{eq22}, we choose $\mathbb{X}=L^2([0,\,\pi])$ and the operator $A:D(A)\subset \mathbb{X}\to \mathbb{X}$ is given by $Au=-u''$ with domain $$ D(A)=\{u\in \mathbb{X}: u'\in \mathbb{X}, \,u''\in \mathbb{X},\, u(0)=u(\pi)=0\}. $$ Then $-A$ generates an analytic semigroup $T(\cdot)$ such that $\|T(t)\|\leq e^{-t}$, $t\geq 0$ (\cite{Lunardi}). Moreover, the eigenvalues of $A$ are $n^2\pi^2$ and the corresponding normalized eigenvectors are $e_n(x)=\sqrt{2}\sin(n\pi x),\, n=1,2,\cdots$. Hence, we have \begin{itemize} \item[(a)] $Au= \sum_{n=1}^\infty n^2\pi^2\langle u,e_n\rangle e_n$ if $ u\in D(A)$; \item[(b)] $A^{-1/2}u= \sum_{n=1}^\infty \frac{1}{n}\langle u, e_n\rangle e_n$ if $ u\in \mathbb{X}$; \item[(c)] The operator $A^{1/2}$ is given by $$ A^{1/2}u= \sum_{n=1}^\infty n\langle u,e_n\rangle e_n $$ for each $u\in D(A^{1/2})=\{u\in \mathbb{X}: \sum_{n=1}^\infty \frac{1}{n}\langle u,e_n\rangle e_n\in \mathbb{X}\} $. \end{itemize} Let $\mathbb{X}_{1/2}= \big(D(A^{1/2}),|\cdot|_{1/2}\big)$ where $| x |_{1/2}=\| A^{1/2} x\|_2$ for each $x\in D(A^{1/2})$. Let $\mathcal{C}_\alpha$ be the Banach space $C([-r\,,0],\mathbb{X}_{1/2})$ equipped with norm $|\cdot|_\infty$. We define $g:\mathbb{R}^+\times \mathbb{X}_{1/2}\to \mathbb{X}_{1/2}$ and $\varphi :[-r,0]\times[0,\pi]\to \mathbb{X}_{1/2} $ by $g(t,u(t))(x)=g(t,u(t,x))$ and $\phi(\theta)(x)=\phi(\theta,x)$ respectively. We define the operator $L$ by $$ L(\phi)(x)= \int_{-r}^0 q(\theta)\phi(\theta)(x) d\theta \quad\text{for } x\in [0,\,\pi],\; \phi \in \mathcal{C}_{1/2}. $$ we have $ A^{1/2} \phi(\theta)(x)\in L^2([-r,\,0])$ since $\phi \in \mathcal{C}_{1/2}$ . It follows that \begin{align*} \big| A^{1/2} L(\phi)(x)\big|^2 &\leq \int_{-r}^0q(\theta)^2 d\theta \int_{-r}^0\big| A^{1/2} \phi(\theta)(x)\big|^2d\theta\\ &\leq r \big( {\sup_{ -r\leq \theta\leq 0}} q(\theta)\big)^2 \int_{-r}^0| A^{1/2} \phi(\theta)(x)|^2 d\theta \end{align*} since $q$ is continuous on $[-r,\,0]$ which is a compact set of $\mathbb{R}$. Therefore we deduce that \begin{align*} \int_0^\pi | A^{1/2} L(\phi)(x)|^2 \, dx &\leq r \big( {\sup_{ -r\leq \theta\leq 0}} q(\theta)\big)^2 \int_0^\pi\int_{-r}^0| A^{1/2} \phi(\theta)(x)|^2 d\theta\, dx\\ &= r \big( {\sup_{ -r\leq \theta\leq 0}} q(\theta)\big)^2 \int_{-r}^0 \int_0^\pi | A^{1/2} \phi(\theta)(x)|^2 dx \, d\theta. \end{align*} Hence, we obtain $$ | L(\phi)|_{1/2} \leq r^2 \big( {\sup_{ -r\leq \theta\leq 0}} q(\theta)\big)^2 | \phi|_{\mathcal{C}_{1/2}}^2 . $$ This means that $L$ is a bounded linear operator from $\mathcal{C}_{1/2} $ to $\mathbb{X}_{1/2}$. Therefore, \eqref{syst2} takes the abstract form \eqref{eq22}. Assume $\int_{-r}^0 | q(\theta)| d\theta<1$ and that the function $g:{R}^+ \times \mathbb{X}_{\alpha} \to \mathbb{X}_{\alpha}$, $t \to g(t,u)$ is continuous and there exists a constant $K_f \ge 0$ such that $$ | g(t,u)-g(t,v)|_\alpha \le K_g | u-v |_\alpha \quad\text{for all } t \in \mathbb{R}^+,\; ( u,v )\in \mathbb{X}^2. $$ Note that such a function exists. Take for instance Let $f(t,x)= e^{-t} x$ then $| f(t,x)-f(t,y)|_{1/2} \leq | x-y|_{1/2}$. \begin{theorem} \label{thm7} Assume that $g$ is uniformly S-asymptotically $\omega$-periodic on bounded sets and asymptotically uniformly continuous on bounded sets. Then System \eqref{syst2} has a unique solution defined on $[-r,\infty[$ such that its restriction on $\mathbb{R}^+$ belongs to $ SAP_{\omega}(\mathbb{X}_{\alpha})$ provided $(r^2 \left( {\sup_{ -r\leq \theta\leq 0}} q(\theta)\right)^2 +K_g)<1$. \end{theorem} \begin{proof} It suffices to apply Theorem \ref{thm6}, observing that (H2) is satisfied since $r^2 \big( {\sup_{ -r\leq \theta\leq 0}} q(\theta)\big)^2 +K_g<1$ and $M=\delta=1$. \end{proof} \begin{thebibliography}{00} \bibitem{adimy0} M. Adimy, A. Elazzouzi, K. Ezzinbi; \emph{Reduction principle and dynamic behaviors for a class of partial functional differential equations}, Nonlinear Analysis. Theory, Methods and Applications, 71(56), 2009, 1709-1727. \bibitem{Adimy} M. Adimy, K. Ezzinbi; \emph{Existence and linearized stability for partial neutral functional differential equations}, Differential Equations Dynam. Systems, Vol.7 (1999) 371-417. \bibitem{Adimy0} M. Adimy, K. Ezzinbi, M.Laklach; \emph{Spectral decomposition for partial neutral functional differential equations}, Canadian. Appl. Math. Q. 9, Vol. 1 (2001), 1-34. \bibitem{Blot} J. Blot, P. Cieutat, G. M. N'Gu\'er\'ekata; \emph{S-asymptotically $\omega$-periodic functions and applications to evolution equations} (preprint). \bibitem{Bochner} S. Bochner; \emph{Continuous mappings of almost automorphic and almost periodic functions,} Proc. Nat. Acad. Sci. USA 52(1964), 907-910. \bibitem{Cuevas} C. Cuevas, M. Pierri, A. Sepulveda; \emph{Weighted S-asymptotically $\omega$-periodic solutions of a class of fractional differential equations}, Advances in Difference Equations, Vol. 2011, Art. ID 584874, 13 p. \bibitem{Diagana} T. Diagana, G. M. N'Gu\'er\'ekata; \emph{Almost automorphic solutions to some classes of partial evolution equations}, Applied Math. Letters 20(2007), 462-466. \bibitem{Dign} T. Diagana; \emph{Existence of Pseudo-Almost Automorphic Mild Solutions to some Nonautonomous Partial Evolution Equations}, Advances in Difference Equations 2011 Article ID 895079, 23 pages doi: 10.1155/2011/895079 \bibitem{Hernandez} T. Diagana, E. Hern\'andez, M. Rabello; \emph{Pseudo almost periodic solutions to some non-autonomous neutral functional differential equations with unbounded delay}, Mathematical and Computer Modelling 45 (2007) 1241-1252. \bibitem{Dimbour} W. Dimbour, G. M. N'Gu\'er\'ekata; \emph{S-asymptotically $\omega$-periodicsolutions to some classes of partial evolution equations}(preprint). \bibitem{Elazzouzi} A. Elazzouzi; \emph{$C^{(n)}$-almost periodic and $C^{(n)}$-almost automorphic solutions for a class of partial functional differential equations with finite delay}. \bibitem{Engel} K. J. Engel, R. Nagel; \emph{One Parameter Semigroups of Linear Evolution Equations}, Grad. text in Math., Vol. 194, Springer(2001). \bibitem{Ezzinbi0} K. Ezzinbi, N. Boukli-Hacene; \emph{Weighted pseudo almost periodic solutions for some partial functional differential equationq}, Vol. 71 (2009) 3612-3621. \bibitem{Ezzinbi} K. Ezzinbi, G. M. N'Gu\'er\'ekata; \emph{Almost automorphic solutions for some partial functional differential equations}, Journal of Mathematical Analysis and Applications, Vol. 328 (2007), 344-358. \bibitem{Lunardi} A. Lunardi; \emph{Analytic semigroup and optimal regularity in parabolic problems}, in PNLDE Vol.16, Birkh\"auser Verlag Basel, 1995. \bibitem{Hale} J. K. Hale, V. Lunel; \emph{Introduction to functional Differential Equations, in: Applied Mathematical Sciences}, Vol. 99, Springer-verlag, New-York/Berlin (1993). \bibitem{traweb} C. C. Travis, G. F. Webb; \emph{Existence and stability for partial functional differential equations}, Transactions of the American Mathematical Society 200 (1974), 395-418. \bibitem{Henriquez1} H. R. Henr\'iquez, M. Pierre, P. T\'aboas; \emph{On S-asymptotically $\omega$-periodic function on Banach spaces and applications}, J. Math. Anal. Appl 343(2008), 1119-1130. \bibitem{Henriquez2} H. R. Henr\'iquez, M. Pierre, P. T\'aboas; \emph{Existence of S-asymptotically $\omega$-periodic solutions for abstract neutral equations}, Bull. Austr. Math. Soc 78(2008), 365-382. \bibitem{hsiang} Hsiang Liu, Jung-Chan Chang; \emph{Existence for a class of partial differential equations with nonlocal conditions}, Nonlinear Analysis, TMA, (in press). \bibitem{Lizama} C. Lizama, G. M. N'Gu\'er\'ekata; \emph{Bounded mild solutions for semilinear integro differential equations in Banach spaces}, Integral Equations and Operators Theory, 68(2010), 207-227. \bibitem{Mophou} G. Mophou, G. M. N'Gu\'er\'ekata ; \emph{Mild Solutions for Semilinear Fractional Differential Equations}, Electronic Journal of Differential Equations, Vol.2009(2009), No. 21, pp.1-9. \bibitem{gasston1} G. M. N'Gu\'er\'ekata; \emph{Almost Automorphy and almost periodic function in Abstract Spaces} Kluwer Academic / Plenum Publisher, New York-Berlin-Moscow, 2001. \bibitem{gasston2} G. M. N'Gu\'er\'ekata; \emph{Topics in almost automorphy}, Springer-Verlag, New York 2005. \bibitem{Nicola} S. Nicola, M. Pierri; \emph{A note on S-asymptotically $\omega$-periodic functions}, Nonlinear Analysis, R.W.A., 10 (2009), 2937-2938. \bibitem{Pazy} A. Pazy; \emph{Semigroups of linear Operators and Application to Partial Differential Equations}, in:Applied Mathematical sciences, vol. 44, Springer-Verlag, New-York, 2001. \bibitem{Prato} G. Da Prato, E. Sinestrari; \emph{Differential operators with nondense domains}, Ann. Sci. Norm. Super. Pisa Cl. Sci., Vol. 2 ( 1987) 285-344. \bibitem{Travis} C. C. Travis, G. F. Webb; \emph{ Partial differential equations with deviating arguments in the time variable}, Journal of Mathematical Analysis and Applications, 56 (1976) 397-409. \bibitem{Wu} J. Wu; \emph{Theory and Applications of Partial Functional Differential Equations}, Springer (1996). \end{thebibliography} \end{document}