Abstract
In this article, we study the set-valued dynamics related to some Euler-Lagrange type functional equations of convex-valued m-mappings. We deal with perturbations of these equations. In order to do this, we use the Banach contraction principle and the Hausdorff distance. Several outcomes on approximate solutions of a few important classic equations are discussed and some applications are given.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
1 Introduction
Let X be a normed space, Y a Banach space and \(\varepsilon > 0\). Smajdor [28] and Gajda and Ger [10] observed that if a mapping \(f: X \rightarrow Y\) satisfies
where \(B(0,\varepsilon )\) is the closed ball of radius \(\varepsilon \) centered at 0, then the set-valued mapping
is subadditive (i.e., \(F(x + y) \subseteq F(x) + F(y)\) for \(x,y\in X\)) and the mapping \(g: X \rightarrow Y\), which satisfies
is an additive selection of F (i.e., \(g(x) \in F(x)\) and \(g(x + y) = g(x) + g(y)\) for \(x,y \in X\)).
There arises the natural question under what conditions a subadditive set-valed function admits an additive selection. An answer to this question can be found in Gajda and Ger [10]. In fact, they proved that if \((S,+)\) is a commutative semigroup with zero element, Y is a Banach space and \(F: S \rightarrow \mathcal {P}(Y)\) is a subadditive mapping with bounded diameter and nonempty, convex and closed values, then F admits a unique additive selection.
It is interesting that, once we have obtained a result of Gajda-Ger type, we can prove the stability of the functional equations corresponding to the functional inclusions considered. For more information, see, e.g., [3,4,5,6, 9, 16,17,18, 21,22,26].
Let f be a mapping between two vector spaces and \(a\ne \pm 1\) be a fixed nonzero integer. For \(m=1,2,3,4\), the functional equation
is equivalent to the additive, quadratic, cubic and quartic functional equations, respectively. For convenience, a solution of the functional equation (1.1) is said to be an m-mapping; see, e.g., [11, 12, 29].
Let \(n > 1\) be an integer, \(M =\{1,\ldots ,n\}\) and \(\mathcal {W}=\{I\subseteq M:1\in I \}\). Denote \(M\setminus I\) by \(I^c\) for \(I\in \mathcal {W}\). In this paper, we consider the following Euler-Lagrange type functional equations
where \(m\in \{1,2,3\}\), f is a mapping between two vector spaces and \(a_{1},\ldots ,a_{n}\) are fixed nonzero integers with \(a_{1}\ne \pm 1\) and \(a_{n}=1\). Using the Banach contraction principle and the Hausdorff distance, we deal with perturbations of set-valued versions of the functional equations (1.2) and (1.3). Some particular cases of our results are discussed. More importantly, the corresponding single-valued functional equations acting as special cases will be included in our results.
Note that if f satisfies (1.3), then f is quartic, see [14, Theorem 2.2] (the single-valued and set-valued versions of (1.3) with \(n=2\), \(a_1=2\) and \(a_2=1\) were studied in [20, 24]). If f satisfies (1.2) with \(m=3\), then f is cubic, see [13, Theorem 2.2] (the single-valued and set-valued versions of (1.2) with \(m=3\), \(n=2\), \(a_1=2\) and \(a_2=1\) were studied in [15, 24, 25]). It is easy to verify from [19, Lemma 3.2] that if f satisfies (1.2) with \(m=2\), then f is quadratic (the equation (1.2) with \(m=2\), \(n=2\), \(a_1=2\) and \(a_2=1\) was studied in [7]). Also, it is easy to verify from [12, Theorem 2.1] that if f satisfies (1.2) with \(m=1\), then f is additive (the equation (1.2) with \(m=1\), \(n=2\), \(a_1=2\) and \(a_2=1\) was studied in [1]).
2 Set-valued dynamics related to functional equations (1.2) and (1.3)
The Banach fixed point theorem [2] (also known as the Banach contraction principle) is an important tool in the theory of metric spaces because it guarantees the existence and uniqueness of fixed points of certain self mappings of metric spaces and provides a constructive method to find those fixed points.
Theorem 2.1
(The Banach contraction principle). Let (X, d) be a complete metric space, and consider a mapping \(\Lambda :X\rightarrow X\), which is strictly contractive, i.e., \(d(\Lambda x,\Lambda y)\le L\hspace{.03cm} d(x,y)\) for all \(x,y\in X\) and some (Lipschitz constant) \(L\in (0,1)\). Then:
- (i):
-
The mapping \(\Lambda \) has one and only one, fixed point \(e=\Lambda (e)\),
- (ii):
-
The fixed point e is globally attractive, i.e., \(\lim _{n\rightarrow \infty } \Lambda ^nx=e\) for any starting point \(x\in X\),
- (iii):
-
The estimation inequality \(d(x,e)\le \frac{1}{1-L}\hspace{.03cm}d\left( x,\Lambda x\right) \) holds for all \(x\in X\).
Let Y be a Banach space. We denote the set of all nonempty subsets of Y by \(\mathcal {P}_{0}(Y)\) and the set of all convex closed bounded members of \(\mathcal {P}_{0}(Y)\) by \(C_{clb}(Y ) \). The number
is said to be the diameter of \(A \in \mathcal {P}_{0}(Y)\).
For \(A,B\in \mathcal {P}_{0}(Y)\) and \(\lambda ,\eta \in \mathbb {R}\), we write \(A + B = \left\{ x+y:~ x \in A,~ y \in B \right\} \) and \(\lambda A = \left\{ \lambda x:~ x \in A \right\} ;\) it is well known that \( \lambda (A+B)=\lambda A+\lambda B\) and \((\lambda +\eta ) A\subseteq \lambda A+\eta A.\) Furthermore, when A is convex and \(\lambda \eta \ge 0\), we obtain \((\lambda +\eta ) A= \lambda A+\eta A\).
For convex closed elements \( A_1,\ldots ,A_n\in \mathcal {P}_{0}(Y)\), we define
where \(\overline{A_1+\cdots +A_n}\) denotes the closure of \(A_1+\cdots +A_n\).
For any closed bounded elements \(A,A' \in \mathcal {P}_{0}(Y )\), the Hausdorff distance h between A and \( A' \) is defined by
where \({B}_{Y} \) is the closed unit ball in Y. The following result is proved from the definition of the Hausdorff distance and can be found in [8].
Proposition 2.2
Let \( A,A',B,B',C \in C_{clb}(Y ) \) and \( \lambda >0 \). Then
- (i):
-
\(h(A\oplus A',B \oplus B') \le h(A, B) + h(A', B') \);
- (ii):
-
\( h(\lambda A,\lambda B ) =\lambda h(A, B) \);
- (iii):
-
\( h(A\oplus C,B \oplus C) =h(A, B) \).
Let \( (C_{clb}(Y),\oplus , h) \) be endowed with the Hausdorff distance h. In [8, Chapter II], it was proved that \((C_{clb}(Y ),\oplus , h)\) is a complete metric semigroup. Rådström proved that \( (C_{clb}(Y ),\oplus , h) \) is isometrically embedded in a Banach space, see [27, Theorem 2].
From now on, unless otherwise specified, let \(n > 1\) be an integer, V be a vector space, Y be a Banach space, \(\mathbb {R}_+=(0,\infty )\), \(M =\{1,\ldots ,n\}\), \(\mathcal {W}=\{I\subseteq M:1\in I \}\), \(I^c=M{\setminus } I\) for \(I\in \mathcal {W}\), \(a_{1},\ldots ,a_{n}\) be fixed positive integers with \(a_{1}\ne 1\) and \(a_{n}=1\), \(m\in \{1,2,3\}\) be fixed and \(\xi \in \{-1,1\}\) be fixed.
Let us now consider the following set-valued functional equations related to the Euler-Lagrange type functional equations (1.2) and (1.3).
Let f from V to \(C_{clb}(Y ) \) be a set-valued mapping. For \(m = 1\), 2 and 3, f is called convex-valued additive, quadratic and cubic, respectively, if it satisfies the following set-valued functional equation
for \(x_1,\ldots ,x_n\in V\). For convenience, a solution of (2.1) is said to be a convex-valued m-mapping. Also, f is called convex-valued quartic if it satisfies the following set-valued functional equation
for \(x_1,\ldots ,x_n\in V\).
Theorem 2.3
Let \(\omega _{m}: V^{n}\rightarrow \mathbb {R}_+\) be a function such that
for an \(L\in (0,1)\). If \( f: V \rightarrow (C_{clb}(Y ),\oplus , h)\) is a set-valued mapping satisfying
then there exists a unique convex-valued m-mapping \(c_m^*: V \rightarrow (C_{clb}(Y ), \oplus , h) \) such that
Moreover, if V is a normed space and \( t, \pi \) are positive real numbers such that \( \xi m<\xi t \) and
then \(c_m^*\) is single-valued.
Proof
From (2.4) with \((x_1,\ldots ,x_n)=(x,0,\ldots ,0)\), we obtain
Since the range of f is convex, it follows from the last inequality, (2.3) and Proposition 2.2 that
and
Hence
where \(\xi \in \{-1,1\}\) is fixed.
Let us consider a complete generalized metric space \((\Upsilon , d)\), where
and
Put also
We show that \( \Gamma : \Upsilon \rightarrow \Upsilon \) is a strictly contractive operator with the Lipschitz constant L. To do this, fix \(g,g'\in \Upsilon \), \(x \in V\) and \(C_{g,g'} \in [0,\infty )\) with \(d(g, g') \le C_{g,g'}\). Then
and consequently \(d(\Gamma g, \Gamma g') \le Ld(g, g')\) for \(g,g'\in \Upsilon \), as claimed.
We deduce from (2.7) that
We can now apply Theorem 2.1 to deduce that the sequence \(\{\Gamma ^j f \}_{j\mathbb {N}}\) is convergent in \((\Upsilon , d)\) and its limit
is a unique fixed point of \(\Gamma \). Moreover,
which proves (2.5).
It follows from (2.3), (2.4) and Proposition 2.2 that
Taking the limit as \(n\rightarrow \infty \), we observe that
and thus \(c_m^*\) is a solution of (2.1).
Furthermore, by (2.6), we obtain
taking the limit as \(n\rightarrow \infty \) and using (2.8), we observe that \(c_m^*\) is a singleton set. This completes the proof of this theorem. \(\square \)
Remark 2.4
Theorem 2.3 can be applied to \( \omega _{m}(x_1,\ldots ,x_n):=\zeta \sum ^{n}_{i=1}\Vert x_i \Vert ^{p} \) for \( x_1,\ldots ,x_n \in V\), where V is a normed space and \(\zeta ,p\) are positive real numbers with \( \xi m <\xi p\).
Theorem 2.5
Let \(\varpi : V^{n}\rightarrow \mathbb {R}_+\) be a function such that
for an \(L\in (0,1)\). If \( f: V \rightarrow (C_{clb}(Y ),\oplus , h)\) is a set-valued mapping satisfying \(f(0)=\{0\}\) and
then there exists a unique convex-valued quartic mapping \(c^*: V \rightarrow (C_{clb}(Y ), \oplus , h) \) such that
Moreover, if V is a normed space and \( t, \pi \) are positive real numbers such that \( 4\xi <\xi t \) and \( \text {diam}\left( f(x)\right) \le \pi \Vert x\Vert ^{t} \) for \(x\in V\), then \(c^*\) is single-valued.
Proof
Setting \((x_1,\ldots ,x_n)=(x,0,\ldots ,0)\) in (2.10) and using (2.9), the convexity of the range of f and Proposition 2.2, we have
Let us consider a complete generalized metric space \((\Upsilon , d)\), where
and
Put also
We deduce from (2.11) that
The rest of the proof is similar to the proof of Theorem 2.3. \(\square \)
Remark 2.6
Theorem 2.5 can be applied to \( \varpi (x_1,\ldots ,x_n):=\zeta \sum ^{n}_{i=1}\Vert x_i \Vert ^{p} \) for \( x_1,\ldots ,x_n \in V\), where V is a normed space and \(\zeta ,p\) are positive real numbers with \( 4\xi <\xi p\).
Data availability
This article has no additional data.
References
Bae, J.H., Park, W.G.: A functional equation having monomials as solutions. Appl. Math. Comput. 216, 87–94 (2010)
Banach, S.: Sur les opérations dans les ensembles abstraits et leur application aux équations intégrales. Fund. Math. 3, 133–181 (1922)
Brzdȩk, J., Piszczek, M.: Fixed points of some nonlinear operators in spaces of multifunctions and the Ulam stability. J. Fixed Point Theory Appl. 19, 2441–2448 (2017)
Brzdȩk, J., Piszczek, M.: Selections of set-valued maps satisfying some inclusions and the Hyers-Ulam stability. In: Handbook of Functional Equations, Springer Optim. Appl., vol. 96, pp. 83–100. Springer, New York (2014)
Brzdȩk, J., Piszczek, M.: Ulam stability of some functional inclusions for multi-valued mappings. Filomat 31, 5489–5495 (2017)
Brzdȩk, J., Popa, D., Xu, B.: Selections of set-valued maps satisfying a linear inclusion in a single variable. Nonlinear Anal. 74, 324–330 (2011)
Chang, I.S., Kim, H.M.: On the Hyers-Ulam stability of quadratic functional equations. J. Ineq. Pure Appl. Math. 3, 33 (2002)
Castaing, C., Valadier, M.: Convex Analysis and Measurable Multifunctions. Lecture Notes in Mathematics, vol. 580. Springer, Berlin (1977)
Chu, H.Y., Kim, A., Yoo, S.K.: On the stability of the generalized cubic set-valued functional equation. Appl. Math. Lett. 37, 7–14 (2014)
Gajda, Z., Ger, R.: Subadditive multifunctions and Hyers-Ulam stability. Numer. Math. 80, 281–291 (1987)
Ghazanfari, A.G., Alizadeh, Z.: On approximate ternary \(m\)-derivations and \(\sigma \)-homomorphisms. J. Fixed Point Theory 17, 625–640 (2015)
Gordji, M.E., Alizadeh, Z., Khodaei, H., Park, C.: On approximate homomorphisms: a fixed point approach. Math. Sci. 6, 59 (2012)
Gordji, M.E., Khodaei, H., Najati, A.: Approximation of cubic mappings with \(n\)-variables in \(\beta \)-normed left Banach modules on Banach algebras. Bull. Korean Math. Soc. 48, 1063–1078 (2011)
Gordji, M.E., Khodaei, H., Najati, A.: Fixed points and quartic functional equations in \(\beta \)-Banach modules. Results Math. 62, 137–155 (2012)
Jun, K.W., Kim, H.M.: The generalized Hyers-Ulam-Rassias stability of a cubic functional equation. J. Math. Anal. Appl. 274, 867–878 (2002)
Khodaei, H.: On the stability of additive, quadratic, cubic and quartic set-valued functional equations. Results Math. 68, 1–10 (2015)
Khodaei, H.: Selections of generalized convex set-valued functions satisfying some inclusions. J. Math. Anal. Appl. 474, 1104–1115 (2019)
Khodaei, H., Rassias, T.M.: Set-valued dynamics related to generalized Euler-Lagrange functional equations. J. Fixed Point Theory Appl. 20, 32 (2018)
Lee, Y., Chung, S.: Stability for quadratic functional equation in the spaces of generalized functions. J. Math. Anal. Appl. 336, 101–110 (2007)
Lee, S.H., Im, S.M., Hawng, I.S.: Quartic functional equation. J. Math. Anal. Appl. 307, 387–394 (2005)
Lu, G., Park, C.: Hyers-Ulam stability of additive set-valued functional equations. Appl. Math. Lett. 24, 1312–1316 (2011)
Miahi, M., Mirzaee, F., Khodaei, H.: On convex-valued G-m-monomials with applications in stability theory. RACSAM 115, 76 (2021)
Nikodem, K., Popa, D.: On selections of general linear inclusions. Publ. Math. Debrecen 75, 239–249 (2009)
Park, C.: Fixed point method for set-valued functional equation. J. Fixed Point Theory Appl. 19, 2297–2308 (2017)
Park, C., O’Regan, D., Saadati, R.: Stabiltiy of some set-valued functional equations. Appl. Math. Lett. 24, 1910–1914 (2011)
Popa, D.: Additive selections of \((\alpha, \beta )\)-subadditive set valued maps. Glas. Mat. 36, 11–16 (2001)
Rådström, H.: An embedding theorem for space of convex sets. Proc. Am. Math. Soc. 3, 165–169 (1952)
Smajdor, W.: Superadditive set-valued functions. Glas. Mat. 21, 343–348 (1986)
Wang, Z.: Fuzzy approximate \(m\)-mappings in quasi fuzzy normed spaces. Fuzzy Sets Syst. 406, 82–92 (2021)
Funding
No funding was received for conducting this study.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Funding
No funding was received for conducting this study.
Conflict of interest
The author declares no conflict of interest.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Khodaei, H. Set-valued dynamics related to convex-valued m-mappings. Aequat. Math. (2024). https://doi.org/10.1007/s00010-024-01103-y
Received:
Revised:
Accepted:
Published:
DOI: https://doi.org/10.1007/s00010-024-01103-y