On the asymptotic equivalence of a differential system with maxima

Abstract

In this paper, we give some general results on the asymptotic relationship between the solutions of a linear differential system and its perturbed differential system with maxima. Also, we present an example to illustrate our results.

Authors

D. Otrocol
(Tiberiu Popoviciu Institute of Numerical Analysis, Romanian Academy,

Keywords

Differential equations with maxima; Dhage iteration method; hybrid fixed point theorem; approximation of solutions

Cite this paper as:

D. Otrocol, On the asymptotic equivalence of a differential system with maxima, Rend. Circ. Mat. Palermo (2), Vol. 65(2016) no. 3, pp. 387-393.

PDF

2016-Otrocol-On the asymptotic.pdf ??

About this paper

Journal

Rendiconti del Circolo Matematico di Palermo

Publisher Name

Springer-Verlag, Italia

Print ISSN

0009-725X

Online ISSN

1973-4409

MR

MR3571317

ZBL

Google Scholar

[1] Bainov, D.D., Hristova, S., Differential equations with maxima, Pure and applied mathematics. Chapman & Hall/CRC (2011)Google Scholar

[2] Bainov, D.D., Kazakova, N.G.,  A finite difference method for solving the periodic problem for autonomous differential equations with maxima. Math. J. Toyama Univ. 15, 1–13 (1992)MathSciNetzbMATHGoogle Scholar

[3] Besekerski, V.A., Popov, E.P., Theory of automatic regulation system. Nauka, Moskov (1975). (in Russian)Google Scholar

[4] Diamandescu, A., Note on the ψψ-boundedness of the solutions of a system of differential equations. Acta Math. Univ. Comenianae 73(2), 223–233 (2004)MathSciNetzbMATHGoogle Scholar

[5] Georgiev, L., Angelov, V.G., On the existence and uniqueness of solutions for maximum equations. Glasnik Matematički 37(2), 275–281 (2002)MathSciNetzbMATHGoogle Scholar

[6] Gonzáles, P., Pinto, M., Component-wise conditions for the asymptotic equivalence for nonlinear differential equations with maxima. Dyn. Syst. Appl. 20, 439–454 (2011)zbMATHGoogle Scholar

[7] Otrocol, D., Rus, I.A., Functional-differential equations with “maxima” via weakly Picard operators theory. Bull. Math. Soc. Sci. Math. Roumanie 51(99), 253–261 (2008)Google Scholar

[8] Otrocol, D., Rus, I.A., Functional-differential equations with maxima of mixed type argument. Fixed Point Theory 9(1), 207–220 (2008)MathSciNetzbMATHGoogle Scholar

[9] Otrocol, D., Properties of the solutions of system of differential equations with maxima, via weakly Picard operator theory. Commun. Appl. Anal. 17(1), 99–107 (2013)MathSciNetzbMATHGoogle Scholar

[10] Otrocol, D., Systems of functional differential equations with maxima, of mixed type. Electron. J. Qual. Theory Differ. Equ. 2014(5), 1–9 (2014)Google Scholar

[11] Otrocol, D., Ilea, V.A., Qualitative properties of functional differential equation. Electron. J. Qual. Theory Differ. Equ. 2014(47), 1–8 (2014)Google Scholar

[12] Olaru, M., Olaru, V.,  CgCg asymptotic equivalence for some functional equation of type Volterra. Gen. Math. 14(1), 31–40 (2006)MathSciNetzbMATHGoogle Scholar

[13] Piccinini, L.C., Stampacchia, G., Vidossich, G., Ordinary differential equations in RnRn. Springer-Verlag (1984)Google Scholar

[14] Stepanov, E., On solvability of some boundary value problems for differential equations with ”maxima”. Topol. Methods Nonlinear Anal. 8, 315–326 (1996)Google Scholar

[15] Talpalaru, P., On stability of non-linear differential systems. Bul. Inst. Pol. Iaşi 27(1–2), 43–48 (1977)MathSciNetzbMATHGoogle Scholar

On the asymptotic equivalence of a differential system with maxima

Diana Otrocol
Abstract

In this paper, we give some general results on the asymptotic relationship between the solutions of a linear differential system and its perturbed differential system with maxima. Also, we present an example to illustrate our results.

T. Popoviciu Institute of Numerical Analysis, Romanian Academy, Cluj-Napoca, Romania

MSC 2010: 34K05, 34C41, 47H10.
Keywords: Differential equations with maxima, asymptotic equivalence.

1 Introduction

In the last few decades, great attention has been paid to automatic control systems and their applications to computational mathematics and modeling. Many problems in control theory correspond to the maximal deviation of the regulated quantity. A classical example is that of an electric generator. In this case, the mechanism becomes active when the maximum voltage variation that is permitted is reached in an interval of time It=[th,t]I_{t}=[t-h,t] with hh a positive constant. The equation which describes the action of this regulator has the form

V(t)=δV(t)+pmaxsItV(s)+F(t),V^{\prime}(t)=-\delta V(t)+p\underset{s\in I_{t}}{\mathrm{max}}V(s)+F(t),

where δ\delta and pp are constants that are determined by characteristic of system, V(t)V(t) is the voltage and F(t)F(t) is the effect of the perturbation that appears associated to the change of voltage [2]. Recently, there is an increasing interest towards equations which contain maxima. We mention the work in [1], [2], [5], [7]-[9], [14].

For some results concerning the asymptotic behavior and asymptotic equivalence of systems of differential equations, we refer in particular to Diamandescu [4], Gonzales and Pinto [6], Olaru [12], Piccinini et.al. [13], Talpalaru [15] and the references cited therein.

Consider the following differential system

x(t)=A(t)x(t),tax^{\prime}(t)=A(t)x(t),\ t\geq a (1.1)

and the perturbed differential system with maxima

y(t)=A(t)y(t)+f(t,y(t),maxξ[a,t]y(ξ)),ta.y^{\prime}(t)=A(t)y(t)+f(t,y(t),\underset{\xi\in[a,t]}{\max}y(\xi)),\ t\geq a. (1.2)

In [9], the author proved existence and uniqueness, inequalities of Čaplygin type and data dependence (monotony, continuity) results for the solution of the Cauchy problem associated to the system (1.2), using weakly Picard operator technique. In this paper the hypotheses which will be imposed upon our equations are extensions of those suggested in [13]. Our results extend and improve corresponding theorems in the existing literature (see, e.g. [7]-[12]). Also, in the end we present an example to illustrate our results.

2 Definitions, notations and hypotheses

Let n\mathbb{R}^{n} be the Euclidian nn-space. For u=(u1,,un)Tnu=(u_{1},\ldots,u_{n})^{T}\in\mathbb{R}^{n}, let u:=max{|u1|,,|un|}\left\|u\right\|:=\max\{\left|u_{1}\right|,\ldots,\left|u_{n}\right|\} be the norm of uu. For a matrix AMn×n(),A=(aij),A\in M_{n\times n}(\mathbb{R}),\ A=(a_{ij}), we define the norm |A|\left|A\right| of AA by |A|:=supu1Au\left|A\right|:=\underset{\left\|u\right\|\leq 1}{\sup}\left\|Au\right\|. So,

|A|=max1inj=1n|aij|.\left|A\right|=\underset{1\leq i\leq n}{\max}\sum\limits_{j=1}^{n}\left|a_{ij}\right|.

In the above equations we consider that y0ny_{0}\in\mathbb{R}^{n}, fC([a,[×n×n,n)f\in C([a,\infty[\times\mathbb{R}^{n}\times\mathbb{R}^{n},\mathbb{R}^{n}) and let BC([a,[,n):={yC([a,[,n)|yBC([a,\infty[,\mathbb{R}^{n}):=\{y\in C([a,\infty[,\mathbb{R}^{n})|\ y bounded with y:=maxt[a,]{|y1|,,|yn|}\left\|y\right\|:=\underset{t\in[a,\infty]}{\max}\left\{\left|y_{1}\right|,\ldots,\left|y_{n}\right|\right\}.

Let X1X_{1} denote the subspace of n\mathbb{R}^{n} consisting of all vectors which are values of bounded solutions of (1.1) for t=0t=0 and let X2 X_{2\text{ }} an arbitrary closed fixed subspace of n\mathbb{R}^{n}, supplementary to X1.X_{1}. We denote by P1,P_{1}, the projection of n\mathbb{R}^{n} on X1X_{1}, (that is P1P_{1} is a bounded linear operator, P1:nn,P12=P1,P_{1}:\mathbb{R}^{n}\rightarrow\mathbb{R}^{n},\ P_{1}^{2}=P_{1},\ KerP1=X2P_{1}=X_{2}) and P2=IP1P_{2}=I-P_{1} the projection onto X2X_{2}.

A solution of the equation (1.1) is a function xC1([a,[,n)x\in C^{1}([a,\infty[,\mathbb{R}^{n}) which satisfy the equation. Now, we give the definition of asymptotic equivalence that we need in the sequel.

Definition 2.1

The equations (1.1) and (1.2) are asymptotically equivalent if for every solution xx of (1.1), there is a solution yy of (1.2) such that

limtx(t)y(t)=0,\underset{t\rightarrow\infty}{\lim}\left\|x(t)-y(t)\right\|=0, (2.1)

and conversely, for each solution yy of (1.2) there is a solution xx of (1.1) such that (2.1) holds.

We consider the following differential equation with maxima

y(t)=A(t)y(t)+f(t,y(t),maxaξty(ξ)),t[a,[,y^{\prime}(t)=A(t)y(t)+f(t,y(t),\underset{a\leq\xi\leq t}{\max}y(\xi)),\ t\in[a,\infty[, (2.2)

with the condition

y(a)=y0.y(a)=y_{0}. (2.3)

We suppose that the following condition holds:

  • (C)

    there exists Lf:[a,[nL_{f}:[a,\infty[\rightarrow\mathbb{R}^{n} with aLf(s)𝑑s<\int_{a}^{\infty}L_{f}(s)ds<\infty such that

    f(t,u1,v1)f(t,u2,v2)Lf(t)max{|u1u2|,|v1v2|},\left\|f(t,u_{1},v_{1})-f(t,u_{2},v_{2})\right\|\leq L_{f}(t)\max\{\left|u_{1}-u_{2}\right|,\left|v_{1}-v_{2}\right|\},

    t[a,[\forall t\in[a,\infty[ and ui,vin,i=1,2u_{i},v_{i}\in\mathbb{R}^{n},i=1,2.

Throughout this paper we note by X(t)X(t) the fundamental matrix of the system (1.1). We remark that if yC1([a,[,n)y\in C^{1}([a,\infty[,\mathbb{R}^{n}) is a solution of the problem (2.2)–(2.3), then yy is a solution of

y(t)=X(t)X1(a)y0+atX(t)X1(s)f(s,y(s),maxaξsy(ξ))ds,t[a,[y(t)=X(t)X^{-1}(a)y_{0}+\int_{a}^{t}X(t)X^{-1}(s)f(s,y(s),\underset{a\leq\xi\leq s}{\max}y(\xi))ds,t\in[a,\infty[ (2.4)

and if yC([a,[,n)y\in C([a,\infty[,\mathbb{R}^{n}) is a solution of (2.4), then yC1([a,[,n)y\in C^{1}([a,\infty[,\mathbb{R}^{n}) and is a solution of (2.2)–(2.3).

Also, if yC1([a,[,n)y\in C^{1}([a,\infty[,\mathbb{R}^{n}) is a solution of the problem (2.2), then yy is a solution of

y(t)=X(t)X1(a)y(a)+atX(t)X1(s)f(s,y(s),maxaξsy(ξ))ds,t[a,[y(t)=X(t)X^{-1}(a)y(a)+\int_{a}^{t}X(t)X^{-1}(s)f(s,y(s),\underset{a\leq\xi\leq s}{\max}y(\xi))ds,t\in[a,\infty[ (2.5)

and if yC([a,[,n)y\in C([a,\infty[,\mathbb{R}^{n}) is a solution of (2.5) then yC1([a,[,n)y\in C^{1}([a,\infty[,\mathbb{R}^{n}) and is a solution of (2.2).

Relative to problem (2.2)–(2.3) we have the following existence and uniqueness theorem (see [9], Theorem 3.1).

Theorem 2.2

We suppose that:

  • (i)

    the condition (C)(C) holds;

  • (ii)

    X(t)X1(s)K\left\|X(t)X^{-1}(s)\right\|\leq K for ast<;a\leq s\leq t<\infty;

  • (iii)

    atf(s,0,0)𝑑s<;\int_{a}^{t}\left\|f(s,0,0)\right\|ds<\infty;

  • (iv)

    KatLf(s)𝑑s<1.K\int_{a}^{t}L_{f}(s)ds<1.

Then the problem (2.2)–(2.3) has, in BC([a,[,n)BC([a,\infty[,\mathbb{R}^{n}), a unique solution and this solution is the uniform limit of the successive approximations.

3 Main result

In this section we show that, under some conditions, for a given solution xx of (1.1) there exists a solution yy of (1.2) such that (2.1) holds. We consider the case when P1+P2=1nP_{1}+P_{2}=1_{\mathbb{R}^{n}}.

Theorem 3.1

Let X(t)X(t) be a fundamental matrix of the system (1.1). We suppose that

  • (i)

    the condition (C)(C) holds;

  • (ii)

    af(s,0,0)𝑑s<;\int_{a}^{\infty}\left\|f(s,0,0)\right\|ds<\infty;

  • (iii)

    there exists two projectors P1,P2P_{1},P_{2} of n\mathbb{R}^{n} and a constant K>0K>0 such that

    X(t)P1X1(s)K for ast,\displaystyle\left\|X(t)P_{1}X^{-1}(s)\right\|\leq K\text{ for }a\leq s\leq t,
    X(t)P2X1(s)K for ats,\displaystyle\left\|X(t)P_{2}X^{-1}(s)\right\|\leq K\text{ for }a\leq t\leq s,
    limtX(t)P1=0;\displaystyle\underset{t\rightarrow\infty}{\lim}X(t)P_{1}=0;
  • (iv)

    2KaLf(s)𝑑s<12K\int\nolimits_{a}^{\infty}L_{f}(s)ds<1.

Then (1.1) and (1.2) are asymptotically equivalent.

Proof. Let xx be a bounded solution of (1.1). Corresponding to xx we consider the operator

Ty(t)=\displaystyle Ty(t)= x(t)+atX(t)P1X1(s)f(s,y(s),maxξ[a,s]y(ξ))𝑑s\displaystyle x(t)+\int\nolimits_{a}^{t}X(t)P_{1}X^{-1}(s)f(s,y(s),\underset{\xi\in[a,s]}{\max}y(\xi))ds-
tX(t)P2X1(s)f(s,y(s),maxξ[a,s]y(ξ))𝑑s.\displaystyle-\int\nolimits_{t}^{\infty}X(t)P_{2}X^{-1}(s)f(s,y(s),\underset{\xi\in[a,s]}{\max}y(\xi))ds.

We show that BC([a,[,n)BC([a,\infty[,\mathbb{R}^{n}) is invariant for the operator TT.

If yBC([a,[,n)y\in BC([a,\infty[,\mathbb{R}^{n}), then [7]

|f(t,y(t),maxξ[a,t]y(ξ))|\displaystyle\left|f(t,y(t),\underset{\xi\in[a,t]}{\max}y(\xi))\right| |f(t,y(t),maxξ[a,t]y(ξ))f(t,0,0)|+|f(t,0,0)|\displaystyle\leq\left|f(t,y(t),\underset{\xi\in[a,t]}{\max}y(\xi))-f(t,0,0)\right|+\left|f(t,0,0)\right|
Lf(t)max(|y(t)|,|maxξ[a,t]y(t)|)+|f(t,0,0)|\displaystyle\leq L_{f}(t)\max\left(\left|y(t)\right|,\left|\underset{\xi\in[a,t]}{\max}y(t)\right|\right)+\left|f(t,0,0)\right|
Lf(t)y+|f(t,0,0)|.\displaystyle\leq L_{f}(t)\left\|y\right\|+\left|f(t,0,0)\right|.

Let xx be a bounded solution of (1.1) and yBC([a,[,n)y\in BC([a,\infty[,\mathbb{R}^{n}). Then

|Ty(t)|\displaystyle\left|Ty(t)\right|\leq |x(t)|+at|X(t)P1X1(s)f(s,y(s),maxξ[a,s]y(ξ))|𝑑s+\displaystyle\left|x(t)\right|+\int\nolimits_{a}^{t}\left|X(t)P_{1}X^{-1}(s)f(s,y(s),\underset{\xi\in[a,s]}{\max}y(\xi))\right|ds+
+t|X(t)P2X1(s)f(s,y(s),maxξ[a,s]y(ξ))|𝑑s\displaystyle+\int\nolimits_{t}^{\infty}\left|X(t)P_{2}X^{-1}(s)f(s,y(s),\underset{\xi\in[a,s]}{\max}y(\xi))\right|ds
\displaystyle\leq |x(t)|+KyatLf(s)𝑑s+Kat|f(s,0,0)|𝑑s\displaystyle\left|x(t)\right|+K\left\|y\right\|\int_{a}^{t}L_{f}(s)ds+K\int_{a}^{t}\left|f(s,0,0)\right|ds
+KytLf(s)𝑑s+Kt|f(s,0,0)|𝑑s\displaystyle+K\left\|y\right\|\int_{t}^{\infty}L_{f}(s)ds+K\int_{t}^{\infty}\left|f(s,0,0)\right|ds
=\displaystyle= |x(t)|+KyaLf(s)𝑑s+Ka|f(s,0,0)|𝑑s.\displaystyle\left|x(t)\right|+K\left\|y\right\|\int_{a}^{\infty}L_{f}(s)ds+K\int_{a}^{\infty}\left|f(s,0,0)\right|ds.

So

Ty(t)x(t)+KyaLf(s)𝑑s+Ka|f(s,0,0)|𝑑s.\left\|Ty(t)\right\|\leq\left\|x(t)\right\|+K\left\|y\right\|\int_{a}^{\infty}L_{f}(s)ds+K\int_{a}^{\infty}\left|f(s,0,0)\right|ds.

Now we prove that TT is a contraction on BC([a,[,n)BC([a,\infty[,\mathbb{R}^{n}).

|Ty1(t)Ty2(t)|\displaystyle\left|Ty_{1}(t)-Ty_{2}(t)\right|\leq
\displaystyle\leq at|X(t)P1X1(s)||f(s,y1(s),maxξ[a,s]y1(ξ))f(s,y2(s),maxξ[a,s]y2(ξ))|𝑑s\displaystyle\int\nolimits_{a}^{t}\left|X(t)P_{1}X^{-1}(s)\right|\left|f(s,y_{1}(s),\underset{\xi\in[a,s]}{\max}y_{1}(\xi))-f(s,y_{2}(s),\underset{\xi\in[a,s]}{\max}y_{2}(\xi))\right|ds
+t|X(t)P2X1(s)||f(s,y1(s),maxξ[a,s]y1(ξ))f(s,y2(s),maxξ[a,s]y2(ξ))|𝑑s\displaystyle+\int\nolimits_{t}^{\infty}\left|X(t)P_{2}X^{-1}(s)\right|\left|f(s,y_{1}(s),\underset{\xi\in[a,s]}{\max}y_{1}(\xi))-f(s,y_{2}(s),\underset{\xi\in[a,s]}{\max}y_{2}(\xi))\right|ds
\displaystyle\leq atKLf(s)max(|y1(s)y2(s)|,|maxaξsy1(ξ)maxaξsy2(ξ)|)𝑑s\displaystyle\int\nolimits_{a}^{t}KL_{f}(s)\max\left(\left|y_{1}(s)-y_{2}(s)\right|,\left|\underset{a\leq\xi\leq s}{\max}y_{1}(\xi)-\underset{a\leq\xi\leq s}{\max}y_{2}(\xi)\right|\right)ds
+tKLf(s)max(|y1(s)y2(s)|,|maxaξsy1(ξ)maxaξsy2(ξ)|)𝑑s\displaystyle+\int_{t}^{\infty}KL_{f}(s)\max\left(\left|y_{1}(s)-y_{2}(s)\right|,\left|\underset{a\leq\xi\leq s}{\max}y_{1}(\xi)-\underset{a\leq\xi\leq s}{\max}y_{2}(\xi)\right|\right)ds
\displaystyle\leq atKLf(s)y1y2ds+tKLf(s)y1y2ds,y1,y2BC([a,[,n).\displaystyle\int\nolimits_{a}^{t}KL_{f}(s)\left\|y_{1}-y_{2}\right\|ds+\int_{t}^{\infty}KL_{f}(s)\left\|y_{1}-y_{2}\right\|ds,\ \forall y_{1},y_{2}\in BC([a,\infty[,\mathbb{R}^{n}).

So,

Ty1Ty22K(aLf(s)ds)y1y2,y1,y2BC([a,[,n).\left\|Ty_{1}-Ty_{2}\right\|\leq 2K\left(\int_{a}^{\infty}L_{f}(s)ds\right)\left\|y_{1}-y_{2}\right\|,\ \forall y_{1},y_{2}\in BC([a,\infty[,\mathbb{R}^{n}).

From Banach’s fixed point theorem, there exists a unique solution of equation (1.2). Let yy be a solution of (1.2) correspondent to xx. Then

|x(t)y(t)|\displaystyle\left|x(t)-y(t)\right|
at|X(t)P1X1(s)f(s,y(s),maxξ[a,s]y(ξ))|𝑑s+t|X(t)P2X1(s)f(s,y(s),maxξ[a,s]y(ξ))|𝑑s\displaystyle\leq\int\nolimits_{a}^{t}\!\left|X(t)P_{1}X^{-1}(s)f(s,y(s),\underset{\xi\in[a,s]}{\max}y(\xi))\right|ds\!+\!\int\nolimits_{t}^{\infty}\!\left|X(t)P_{2}X^{-1}(s)f(s,y(s),\underset{\xi\in[a,s]}{\max}y(\xi))\right|ds
at1|X(t)P1X1(s)f(s,y(s),maxξ[a,s]y(ξ))|𝑑s+t1t|X(t)P1X1(s)f(s,y(s),maxξ[a,s]y(ξ))|𝑑s\displaystyle\leq\int_{a}^{t_{1}}\!\left|X(t)P_{1}X^{-1}(s)f(s,y(s),\underset{\xi\in[a,s]}{\max}y(\xi))\right|ds\!+\!\int_{t_{1}}^{t}\!\left|X(t)P_{1}X^{-1}(s)f(s,y(s),\underset{\xi\in[a,s]}{\max}y(\xi))\right|ds
+t|X(t)P2X1(s)f(s,y(s),maxξ[a,s]y(ξ))|𝑑s\displaystyle\quad+\int\nolimits_{t}^{\infty}\left|X(t)P_{2}X^{-1}(s)f(s,y(s),\underset{\xi\in[a,s]}{\max}y(\xi))\right|ds
|X(t)P1|at1|X1(s)||f(s,y(s),maxξ[a,s]y(ξ))|𝑑s+Kt1|f(s,y(s),maxξ[a,s]y(ξ))|𝑑s.\displaystyle\leq\left|X(t)P_{1}\right|\int_{a}^{t_{1}}\left|X^{-1}(s)\right|\left|f(s,y(s),\underset{\xi\in[a,s]}{\max}y(\xi))\right|ds+K\int_{t_{1}}^{\infty}\left|f(s,y(s),\underset{\xi\in[a,s]}{\max}y(\xi))\right|ds.

We fix ε>0\varepsilon>0 and determine t1at_{1}\geq a such that the second integral in the right-hand side of the above inequality is less than ε/2\varepsilon/2. With t1t_{1} fixed, from hypothesis (iii), we have that the first term on the right-hand side of the above inequality tends to zero as xx tends to infinity. We may therefore conclude that limtx(t)y(t)=0\underset{t\rightarrow\infty}{\lim}\left\|x(t)-y(t)\right\|=0.

Similarly, we take a bounded solution yy of (1.2) and we observe that

x(t)\displaystyle x(t) =y(t)atX(t)P1X1(s)f(s,y(s),maxξ[a,s]y(ξ))𝑑s\displaystyle=y(t)-\int\nolimits_{a}^{t}X(t)P_{1}X^{-1}(s)f(s,y(s),\underset{\xi\in[a,s]}{\max}y(\xi))ds
+tX(t)P2X1(s)f(s,y(s),maxξ[a,s]y(ξ))𝑑s\displaystyle\quad+\int\nolimits_{t}^{\infty}X(t)P_{2}X^{-1}(s)f(s,y(s),\underset{\xi\in[a,s]}{\max}y(\xi))ds

satisfies equation (1.1) and (2.1) holds.   

For the case when P1=P2=1nP_{1}=P_{2}=1_{\mathbb{R}^{n}}, we have the following result.

Theorem 3.2

Let X(t)X(t) be a fundamental matrix of the system (1.1). We suppose that

  • (i)

    the condition (C)(C) holds;

  • (ii)

    af(s,0,0)𝑑s<;\int_{a}^{\infty}\left\|f(s,0,0)\right\|ds<\infty;

  • (iii)

    there exists a constant K>0K>0 such that

    X(t)X1(s)K for ast<;\displaystyle\left\|X(t)X^{-1}(s)\right\|\leq K\text{ for }a\leq s\leq t<\infty;
    limtX(t)=0\displaystyle\underset{t\rightarrow\infty}{\lim}X(t)=0
  • (iv)

    2KaLf(s)𝑑s<12K\int\nolimits_{a}^{\infty}L_{f}(s)ds<1.

Then (1.1) and (1.2) are asymptotically equivalent.

Proof. The proof follow the same steps as in Theorem 3.1. Let xx be a bounded solution of (1.1). Corresponding to xx we consider the operator

Ty(t)=\displaystyle Ty(t)= x(t)+ctX(t)X1(s)f(s,y(s),maxξ[a,s]y(ξ))𝑑s\displaystyle x(t)+\int\nolimits_{c}^{t}X(t)X^{-1}(s)f(s,y(s),\underset{\xi\in[a,s]}{\max}y(\xi))ds-
tX(t)X1(s)f(s,y(s),maxξ[a,s]y(ξ))𝑑s.\displaystyle-\int\nolimits_{t}^{\infty}X(t)X^{-1}(s)f(s,y(s),\underset{\xi\in[a,s]}{\max}y(\xi))ds.

We show that BC([a,[,n)BC([a,\infty[,\mathbb{R}^{n}) is invariant for the operator TT and we prove that TT is a contraction on BC([a,[,n).BC([a,\infty[,\mathbb{R}^{n}).

Ty1Ty22K(aLf(s)ds)y1y2,y1,y2BC([a,[,n).\left\|Ty_{1}-Ty_{2}\right\|_{\infty}\leq 2K\left(\int_{a}^{\infty}L_{f}(s)ds\right)\left\|y_{1}-y_{2}\right\|_{\infty},\ \forall y_{1},y_{2}\in BC([a,\infty[,\mathbb{R}^{n}).

From Banach’s fixed point theorem, there exists a unique solution of equation (1.2). Let yy be a solution of (1.2) correspondent to xx. Then

|x(t)y(t)|\displaystyle\left|x(t)-y(t)\right|\leq
|X(t)|at1|X1(s)||f(s,y(s),maxξ[a,s]y(ξ))|𝑑s+Kt1|f(s,y(s),maxξ[a,s]y(ξ))|𝑑s.\displaystyle\leq\left|X(t)\right|\int_{a}^{t_{1}}\left|X^{-1}(s)\right|\left|f(s,y(s),\underset{\xi\in[a,s]}{\max}y(\xi))\right|ds+K\int_{t_{1}}^{\infty}\left|f(s,y(s),\underset{\xi\in[a,s]}{\max}y(\xi))\right|ds.

We fix ε>0\varepsilon>0 and determine t1at_{1}\geq a such that the second integral in the right-hand side of the above inequality is less than ε/2\varepsilon/2. With t1t_{1} fixed, from hypothesis (iii), we have that the first term on the right-hand side of the above inequality tends to zero as xx tends to infinity. We may therefore conclude that limtx(t)y(t)=0\underset{t\rightarrow\infty}{\lim}\left\|x(t)-y(t)\right\|=0, and the theorem is proved.

Similarly, we take a bounded solution yy of (1.2) and we observe that

x(t)\displaystyle x(t) =y(t)atX(t)X1(s)f(s,y(s),maxξ[a,s]y(ξ))𝑑s\displaystyle=y(t)-\int\nolimits_{a}^{t}X(t)X^{-1}(s)f(s,y(s),\underset{\xi\in[a,s]}{\max}y(\xi))ds
+tX(t)X1(s)f(s,y(s),maxξ[a,s]y(ξ))𝑑s\displaystyle\quad+\int\nolimits_{t}^{\infty}X(t)X^{-1}(s)f(s,y(s),\underset{\xi\in[a,s]}{\max}y(\xi))ds

satisfies equation (1.1) and (2.1) holds.   

4 Application

In what follows we consider a differential system with maxima

LI(t)=M(t)I(t)+Nmaxs[0,t]I(s)+kI3(t)1+I2(t),t0,LI^{\prime}(t)=M(t)I(t)+N\underset{s\in[0,t]}{\max}I(s)+k\frac{I^{3}(t)}{1+I^{2}(t)},\ t\geq 0,

where the unknown function I(t)I(t) is the electric current, L0,MM2×2(),NL\neq 0,\ M\in M_{2\times 2}(\mathbb{R}),N and kk are constants. It is derived treating the original automatic regulation phenomenon ([3]) without linearization. Then we can formulate an initial-value problem for the above system as follows:

{I(t)=L1M(t)I(t)+L1f(t,I(t),maxs[0,t]I(s)),t0,I(t)=y0,y02\left\{\begin{array}[c]{l}I^{\prime}(t)=L^{-1}M(t)I(t)+L^{-1}f(t,I(t),\underset{s\in[0,t]}{\max}I(s)),\ t\geq 0,\\ I(t)=y_{0},\ y_{0}\in\mathbb{R}^{2}\end{array}\right. (4.1)

where M(t)M(t) is the fundamental matrix of the nonperturbed system

J(t)=L1M(t)J(t).J^{\prime}(t)=L^{-1}M(t)J(t). (4.2)

The functionf(t,u,v)=L1(Nv+ku31+u2)\ f(t,u,v)=L^{-1}\left(Nv+k\frac{u^{3}}{1+u^{2}}\right) satisfies condition (C)(C), i.e.,

f(t,u1,v1)f(t,u2,v2)|L1|max{c|k||u1u2|,|N||v1v2|},\left\|f(t,u_{1},v_{1})-f(t,u_{2},v_{2})\right\|\leq\left|L^{-1}\right|\max\{c\left|k\right|\left|u_{1}-u_{2}\right|,\left|N\right|\left|v_{1}-v_{2}\right|\},

with cc=const. The condition (iii)(iii) of Theorems 3.1 and 3.2 is satisfied with P1=(1000)P_{1}=\left(\begin{array}[c]{cc}1&0\\ 0&0\end{array}\right) and P2=(1000)P_{2}=\left(\begin{array}[c]{cc}1&0\\ 0&0\end{array}\right). By Theorems 3.1 and 3.2, for every bounded solution JJ of equation (4.2), there exists a unique bounded solution II of equation (4.1) such that limtJ(t)I(t)=0\underset{t\rightarrow\infty}{\lim}\left\|J(t)-I(t)\right\|=0 holds and conversely, for every bounded solution II of equation (4.1), there exists a unique bounded solution JJ of equation (4.2) such that limtJ(t)I(t)=0\underset{t\rightarrow\infty}{\lim}\left\|J(t)-I(t)\right\|=0 holds.

References

  • [1] D.D. Bainov, S. Hristova, “Differential equations with maxima”, Chapman & Hall/CRC Pure and Applied Mathematics, 2011.
  • [2] D.D. Bainov, N.G. Kazakova, A finite difference method for solving the periodic problem for autonomous differential equations with maxima, Math. J. Toyama Univ., 15, pp. 1–13, 1992.
  • [3] V.A. Besekerski, E.P. Popov, “Theory of automatic regulation system”, Nauka, Moskov, 1975 (in Russian).
  • [4] A. Diamandescu, Note on the ψ\psi-boundedness of the solutions of a system of differential equations, Acta Math. Univ. Comenianae, 73 (2004), No. 2, 223–233.
  • [5] L. Georgiev, V.G. Angelov, On the existence and uniqueness of solutions for maximum equations, Glasnik Matematički, 37 (2002), No. 2, 275–281.
  • [6] P. Gonzáles, M. Pinto, Component-wise conditions for the asymptotic equivalence for nonlinear differential equations with maxima, Dynamic Systems and Applications, 20 (2011), 439–454.
  • [7] D. Otrocol, I.A. Rus, Functional-differential equations with “maxima” via weakly Picard operators theory, Bull. Math. Soc. Sci. Math. Roumanie, 51(99) (2008), No. 3, 253–261.
  • [8] D. Otrocol, I.A. Rus, Functional-differential equations with maxima of mixed type argument, Fixed Point Theory, 9 (2008), No. 1, pp. 207–220.
  • [9] D. Otrocol, Properties of the solutions of system of differential equations with maxima, via weakly Picard operator theory, Communications in Applied Analysis, 17 (2013), No. 1, 99–107.
  • [10] D. Otrocol, Systems of functional differential equations with maxima, of mixed type, Electron. J. Qual. Theory Differ. Equ., (2014), No. 5, 1–9.
  • [11] D. Otrocol, V.A. Ilea, Qualitative properties of functional differential equation, Electron. J. Qual. Theory Differ. Equ., (2014), No. 47, 1–8.
  • [12] M. Olaru, V. Olaru, CgC_{g} asymptotic equivalence for some functional equation of type Volterra, General Mathematics, 14 (2006), No. 1, 31–40.
  • [13] L.C. Piccinini, G. Stampacchia, G. Vidossich, “Ordinary differential equations in n\mathbb{R}^{n}”, Springer-Verlag, 1984.
  • [14] E. Stepanov, On solvability of some boundary value problems for differential equations with “maxima”, Topological Methods in Nonlinear Analysis, 8 (1996), 315–326.
  • [15] P. Talpalaru, On stability of non-linear differential systems, Bul. Inst. Pol. Iaşi, 27 (1977), Nos. 1–2, 43–48.

An abbreviated version of the title for the running head: “A differential system with maxima”


“T. Popoviciu” Institute of Numerical Analysis, Romanian Academy,
Fântânele St., No. 57,
Cluj-Napoca, 400320, Romania,
e-mail: dotrocol@ictp.acad.ro.

2016

Related Posts