 Research
 Open access
 Published:
Generalized strong vector quasiequilibrium problems with variable ordering structure
Journal of Inequalities and Applications volumeÂ 2019, ArticleÂ number:Â 116 (2019)
Abstract
In this paper, two kinds of generalized strong vector quasiequilibrium problems with variable ordering structure are considered by using the concept of cosmically upper continuity rather than upper semicontinuity for conevalued mapping. Firstly, a key local property of cosmically upper continuity for conevalued mapping is discussed. Next, under suitable conditions of conecontinuity and coneconvexity for equilibrium mapping, several existence theorems of solutions and closedness of solution sets are established for these two kinds of generalized strong vector quasiequilibrium problems with variable ordering structure. Moreover, an example is given to illustrate the validity of our theorems. These results obtained in this paper extend and develop some recent works in this field.
1 Introduction
The socalled equilibrium problem, which is also called generalized Ky Fan minimax inequality, was firstly studied by Blum and Oettli [1] in 1994 in finitedimensional Euclidean spaces. It provides a brief and unified framework for modeling many problems originating from practice and theory, such as mathematical economics, Nash equilibrium problem, fixed pointed problem, saddle point problem, optimization problem, complementary problem and variational inequality problem, etc. In recent years, equilibrium problems with scalar and vector values objective mappings have been widely studied by researchers, and fruitful results have been established (see, for example, [2,3,4,5,6,7,8,9,10,11,12,13,14] and the references therein). Considering indepth research work, generalized vector quasiequilibrium problems (GVQEP), in which the constraint set is subject to modification, have been put forward and investigated by many people in different settings. From the existing literature, we can see that there are many meaningful works on the existence of solutions for (GVQEP) (see, e.g., [15,16,17,18,19,20,21,22] and the references therein). As applications of these results, they are applied to study the existence of mathematical programs with equilibrium constraints (MPEC), generalized Nash equilibrium, generalized semiinfinite programs, generalized quasivariational inequality, and so on. We would like to emphasize here that the importance of research studies on (GVQEP) is justified by the fact that it is a unified model of other several problems. Furthermore, these studies of (GVQEP) can also be applied to some important reallife problems. For further relevant information, the reader is referred to [23, 24] and the references therein.
On the other hand, considering the preference of a decisionmaker (the partial order in the object space) may vary with time, Yu [25] introduced in 1974 a new type of vector optimization problem with variable ordering structure, where the ordering structure was represented by a conevalued mapping, i.e., by a setvalued mapping where the images are cones. In 1992, Chen and Huang [26] generalized it to variational inequality. Afterwards, there have been many works devoted to the study of various equilibrium, optimization, and variational inequality problems with variable ordering structure (see, for example, [27,28,29,30,31,32] and the references therein). It is worth pointing out that, in these literature sources, the condition of upper semicontinuity for conevalued mapping has been widely used. However, Borde and Crouzeix [33] gave an example to explain that the concept of the upper semicontinuity for setvalued mapping was actually not suitable for the conevalued mapping. Very recently, Eichfelder [34] deeply considered various properties for the conevalued mapping. The author also gave an interesting conclusion, which told us distinctly that the concept of upper semicontinuity for a setvalued mapping is too strict for this special setvalued mapping (conevalued mapping). In order to deal with this problem, Luc and Penot [35] proposed a modified concept of upper semicontinuity called cosmically upper continuity and analyzed its properties. After that, the cosmically upper continuity conditions for conevalued mapping have been also used by some authors to discussed vector optimization problems with variable ordering structure. However, to the best of our knowledge, there is no paper dealing with the existence of solutions for equilibrium problem with variable ordering structure by using the concept of cosmically upper continuity.
Inspired and motivated by the works mentioned above, in this paper, we mainly discuss the existence of solutions for generalized strong vector quasiequilibrium problems with variable ordering structure (for short, GSVQEP) by using the concept of cosmically upper continuity rather than upper semicontinuity for conevalued mapping. Also, we shall discuss the closedness of solution sets for (GSVQEP).
The rest of the paper is organized as follows. In Sect.Â 2, some notations are fixed, several wellknown concepts and mathematical tools are recalled. In Sect.Â 3, a local property for conevalued mapping is discussed under the assumption of cosmically upper continuity. Next, under suitable conditions of conecontinuity and coneconvexity for equilibrium mapping, existence results of solutions and closedness of solution sets are obtained for two kinds of (GSVQEP). Moreover, an example is given to illustrate the validity of our theorem. These results extend and develop some recent works in this field.
2 Preliminaries
Throughout this paper, unless otherwise specified, we always suppose that X and Y are two Hausdorff topological vector spaces and Z is a real normed vector space. Let K be a nonempty, closed, and convex subset of X, E be a nonempty subset of Y. Let \(C : K \rightrightarrows Z\) be a setvalued mapping such that, for all \(x \in K\), \(C(x)\subseteq Z\) is a closed, convex, and pointed cone. We denote by \(\operatorname{conv}(A)\), \(\operatorname{cl}(A)\), and \(\operatorname{cone}(A)\) the convex, closure, and cone hull of the set A, respectively. Since Z is a real normed space, we denote by \(B_{l}=\{x\in Z:\x\< l\}\) and \(S_{l}=\{x\in Z:\x\=l\}\) the open ball around \(0\in Z\) with radius \(l>0\) and its boundary, respectively.
Let \(F : E \times K \times K \rightrightarrows Z\), \(T : K \times K \rightrightarrows E\), \(Q : K \rightrightarrows K\), and \(\widetilde{F}: K \times K \rightrightarrows Z\) be four setvalued mappings with nonempty values. We consider the following two kinds of generalized strong vector quasiequilibrium problems with variable ordering structure (for short, GSVQEP):

(GSVQEP 1)
Find \(\bar{x}\in K\) such that \(\bar{x}\in Q(\bar{x})\), and for all \(y\in Q(\bar{x})\) and \(t\in T(\bar{x},\bar{x})\), one has
$$ F(t,\bar{x},y)\subseteq C(\bar{x}). $$ 
(GSVQEP 2)
Find \(\bar{x}\in K\) such that \(\bar{x}\in Q(\bar{x})\), and for any \(y\in Q(\bar{x})\), there exists some \(t\in T(\bar{x},\bar{x})\) satisfying
$$ F(t,\bar{x},y)\subseteq C(\bar{x}). $$
If \(X=Y\), \(E=K\), \(T(x,y)=\{x\}\), \(Q(x)=K\), and \(F(t,x,y)= \widetilde{F}(x,y)\) for all \(x,y\in K\) and \(t\in E\), then the two kinds of (GSVQEP) collapse to the following strong vector equilibrium problem (for short, SEVP):

(SVEP)
Find \(\bar{x}\in K\) such that
$$ \widetilde{F}(\bar{x},y)\subseteq C(\bar{x}) \quad \mbox{for all } y\in K. $$
In this section, we will give some necessary background on the continuity for setvalued mapping. Following [36], a setvalued mapping \(D: K \rightrightarrows Y\) is said to be upper semicontinuous (for short, u.s.c.) at \(x_{0}\in K\) if, for any open set \(V \subseteq Y\) with \(D(x_{0})\subseteq V\), there exists a neighborhood U of \(x_{0}\) such that \(D(x)\subseteq V\) for all \(x\in U\cap K\). D is said to be lower semicontinuous (for short, l.s.c.) at \(x_{0}\in K\) if, for any open set \(V \subseteq Y\) with \(D(x_{0})\cap V \neq \emptyset \), there exists a neighborhood U of \(x_{0}\) such that \(D(x) \cap V \neq \emptyset \) for all \(x\in U\cap K\). D is said to be u.s.c. (resp. l.s.c.) on K if it is u.s.c. (resp. l.s.c.) at every point \(x\in K\).
Lemma 2.1
Let \(D:K\rightrightarrows Y\) be a setvalued mapping with nonempty values.

(i)
For any given \(x_{0}\in K\), if \(D(x_{0})\) is compact, then D is u.s.c. at \(x_{0}\in K\) if and only if, for any net \(\{x_{\alpha }\}\subseteq K\) with \(x_{\alpha }\to x_{0}\) and for any \(y_{\alpha }\in D(x_{\alpha })\), there exist \(y_{0}\in D(x_{0})\) and a subnet \(\{y_{\beta }\}\) of \(\{y_{\alpha }\}\) such that \(y_{\beta } \to {y_{0}}\).

(ii)
D is l.s.c. at \(x_{0}\in K\) if and only if, for any net \(\{x_{\alpha }\}\) with \(x_{\alpha }\to x_{0}\) and for any \(y_{0} \in D(x_{0})\), there exists a net \(\{y_{\alpha }\}\) with \(y_{\alpha } \in D(x_{\alpha })\) such that \({y_{\alpha }} \to y_{0}\).
In order to deal with conevalued mapping, we need the following modified concept of semicontinuity for conevalued mapping, which is called cosmically upper continuity and was proposed by Luc and Penot in [35]:
A conevalued mapping C is said to be cosmically upper continuous (for short, c.u.c.) at \(x_{0}\in K\) if the mapping \(x\to C(x)\cap \operatorname{cl}(B_{1})\) is u.s.c. at \(x_{0}\in K\). By analogy to the definition of u.s.c., we say that C is c.u.c. on K if it is c.u.c. at every point \(x\in K\).
Remark 2.1
If C is c.u.c. at \(x_{0}\in K\), then âˆ’C is also c.u.c. at \(x_{0}\in K\), and vice versa.
Example 2.1
Considering the following conevalued mapping \(C:\mathbb{R}\rightrightarrows \mathbb{R}^{2}\),
Clearly, \(C(0)=[0,+\infty [\,\times \{0\}\) and \(C(x)\subseteq \mathbb{R}^{2}\) is a closed, convex, and pointed cone for each \(x\in \mathbb{R}\). Moreover, for each \(\delta >0\), we have
Noting that \(C(x)\cap \operatorname{cl}(B_{1})\) is compact, we can conclude easily from the above formula that the mapping \(x\to C(x)\cap \operatorname{cl}(B _{1})\) is u.s.c. at \(x=0\), i.e., C is c.u.c. at \(x=0\).
On the other hand, for any given \(\delta >0\) and \(x\neq 0\), we can take some \(t>0\) such that \(tx>\delta \). Then, by the definition of C, we know \((t,tx)\in C(x)\) but \((t,tx)\notin C(0)+B_{\delta }\). It follows that \(C(x)\nsubseteq C(0)+B_{\delta }\). Hence C is not u.s.c. at \(x=0\).
In the sequel, we also need the following notions related to conecontinuity and coneconvexity of setvalued mappings. Following [38], a setvalued mapping \(D: K \rightrightarrows Z\) is said to be upper \(C(x_{0})\)continuous (resp. lower \(C(x_{0})\)continuous) at \(x_{0}\in K\) if, for any neighborhood V of \(0 \in Z\), there exists a neighborhood U of \(x_{0}\) such that \(D(x)\subseteq D(x_{0}) + V + C(x_{0})\) (resp. \(D(x_{0})\subseteq D(x) + V  C(x_{0})\)) for all \(x\in U\cap K\). D is said to be upper \(C(x)\)continuous (resp. lower \(C(x)\)continuous) on K if it is upper \(C(x)\)continuous (resp. lower \(C(x)\)continuous) at every point \(x \in K\).
Let \(P \subseteq Z\) be a closed, convex, and pointed cone. A setvalued mapping \(D : K \rightrightarrows Z\) is said to be upper properly Pquasiconvex on K if, for any \(u_{1},u_{2}\in K\) and \(t\in [0,1]\), one has
It is said to be lower properly Pquasiconvex on K if, for any \(u_{1},u_{2}\in K\) and \(t\in [0,1]\), one has
Definition 2.1
([39])
Let \(P\subseteq Z\) be a closed, convex, and pointed cone. A nonempty convex subset \(B \subseteq P\) is said to be a base of P if \(0\notin \operatorname{cl}(B)\) and \(P=\operatorname{cone}(B)\).
Lemma 2.2
([39])
For any given closed, convex, and pointed cone \(P\subseteq Z\), if P has a bounded closed base, then \(P^{\Delta }:=\{f\in Z^{*}:\exists a>0, f(x)\geq a\x\, \forall x\in P\} \neq \emptyset \).
Lemma 2.3
([40])
Let \(H : K \rightrightarrows K\) be a setvalued mapping. Suppose that

(i)
for any \(x\in K\), \(x\notin \operatorname{conv}(H(x))\);

(ii)
for each \(y\in K\), \(H^{1}(y)=\{x\in K: y\in H(x)\}\) is open in K;

(iii)
there exist a nonempty and compact subset M of K and a nonempty, compact, and convex subset N of K such that, for each \(x\in K\setminus M\), \(N\cap H(x)\neq \emptyset \).
Then there exists \(x_{0}\in K\) such that \(H(x_{0})=\emptyset \).
3 Main results
In this section, we denote by \(\operatorname{Sol}(F_{1})\), \(\operatorname{Sol}(F_{2})\), and \(\operatorname{Sol}(\widetilde{F})\) the sets of solutions for (GSVQEPÂ 1), (GSVQEP 2), and (SEVP), respectively.
Firstly, we consider a local property for conevalued mapping which is c.u.c.
Lemma 3.1
Let \(x_{0}\in K\). Suppose that \(C(x_{0})\) has a bounded and closed base and C is c.u.c. at \(x_{0}\). Then, for any \(0<\varepsilon <l\), there exists a neighborhood U of \(x_{0}\) such that
Proof
The proof is divided into four steps.
Step 1. For any \(0<\varepsilon <l\) and \(\xi \in C(x_{0})^{ \Delta }\), there exist \(r>\varepsilon \) and \(c>0\) such that
Since \(C(x_{0})\) has a bounded and closed base, it follows from LemmaÂ 2.2 that \(C(x_{0})^{\Delta }\neq \emptyset \). Thus, for any \(\xi \in C(x_{0})^{\Delta }\subseteq Z^{*}\), there exists \(a>0\) such that
Let \(b=\max \{a,\\xi \_{*}\}+1\), where \(\\xi \_{*}\) denotes the norm of the linear operator Î¾. For any \(v\in Z\), it follows from \(b>\\xi \_{*}\) and \(\xi \in Z^{*}\) that
Set \(c=bl+1\) and \(r = (c+a\varepsilon +b\varepsilon )/a\). We can see that \(a(r\varepsilon )b\varepsilon = c >0\) and \(r>\varepsilon \). For any \(v_{0}\in [C(x_{0})+B_{\varepsilon }]\cap S_{r}\), there exist \(v_{1}\in C(x_{0})\) and \(v_{2}\in B_{\varepsilon }\) such that \(v_{0}=v_{1}+v_{2}\). It follows from \(\v_{0}\=r\) and \(v_{2}\in B _{\varepsilon }\) (i.e., \(\v_{2}\<\varepsilon \)) that
Since \(v_{1}\in C(x_{0})\), applying (2) and (4), we have
We conclude from (3) and \(v_{2}\in B_{\varepsilon }\) that \(\xi (v_{2})< b\v_{2}\< b\varepsilon \). This together with (5) implies that
By the arbitrariness of \(v_{0}\), we obtain that (1) is true.
Step 2. \(\operatorname{cone}\{[C(x_{0})+B_{\varepsilon }]\cap S_{r} \}\subseteq \operatorname{cone}\{x\in C(x_{0})+B_{\varepsilon }: \xi (x)=c\} \subseteq \operatorname{cone}\{[C(x_{0})+B_{\varepsilon }]\cap S_{l}\}\).
Let \(C_{1}=\operatorname{cone}\{[C(x_{0})+B_{\varepsilon }]\cap S_{r}\}\) and \(C_{2}=\operatorname{cone}\{x\in C(x_{0})+B_{\varepsilon }: \xi (x)=c\}\). It is easy to see that \(C_{1}\) and \(C_{2}\) are both cones. We claim that \(\{[C(x_{0})+B_{\varepsilon }]\cap S_{r}\}\subseteq C_{2}\). This together with the definition of \(C_{1}\) implies that \(C_{1}\subseteq C_{2}\). In fact, for any \(u_{1}\in [C(x_{0})+B_{\varepsilon }]\cap S _{r}\), it follows from (1) and \(u_{1}\in S_{r}\) that \(\xi (u_{1})>c>0\). Then there exists \(0<\lambda _{1} = c/\xi (u_{1}) <1\) such that
On the other hand, we can see that \(u_{1}\in C(x_{0})+B_{\varepsilon }\) and \(0\in C(x_{0})+B_{\varepsilon }\). This together with the convexity of \(C(x_{0})+B_{\varepsilon }\) and \(0< \lambda _{1} < 1\) implies that
Applying (6), (7) and the definition of \(C_{2}\), we can see that \(\lambda _{1}u_{1}\in C_{2}\). It is obvious that \(u_{1}\in C_{2}\). By the arbitrariness of \(u_{1}\), we can get that \(\{[C(x_{0})+B_{\varepsilon }]\cap S_{r}\}\subseteq C_{2}\). Hence, \(C_{1}\subseteq C_{2}\).
Let \(C_{3}=\operatorname{cone}\{[C(x_{0})+B_{\varepsilon }]\cap S_{l}\}\). Obviously, \(C_{2}\) and \(C_{3}\) are both cones. If we can prove that \(\{x\in C(x_{0})+B_{\varepsilon }: \xi (x)=c\}\subseteq C_{3}\), then we can easily get \(C_{2}\subseteq C_{3}\) by the definition of \(C_{2}\). In fact, for each \(u_{2}\in C(x_{0})+B_{\varepsilon }\) satisfying \(\xi (u_{2})=c\), we claim that \(\u_{2}\\geq l\). If it is not true, then we have \(\xi (u_{2})< b\u_{2}\<bl<c\) from (3), which contradicts \(\xi (u_{2})=c\). Therefore, \(\u_{2}\\geq l\) which implies that there exists \(0<\lambda _{2} \leq 1\) such that
On the other hand, \(u_{2}\in C(x_{0})+B_{\varepsilon }\) and \(0\in C(x_{0})+B_{\varepsilon }\). This together with the convexity of \(C(x_{0})+B_{\varepsilon }\) and \(0< \lambda _{2} < 1\) implies that
Applying (8), (9), and the definition of \(C_{3}\), it is easy to obtain that \(\lambda _{2}u_{2}\in C_{3}\). Obviously, \(u_{2}\in C_{3}\). By the arbitrariness of \(u_{2}\), we have \(\{x\in C(x_{0})+B_{\varepsilon }: \xi (x)=c\}\subseteq C_{3}\), which implies that \(C_{2}\subseteq C_{3}\).
Step 3. There exists a neighborhood U of \(x_{0}\) such that \(C(x)\cap S_{r}\subseteq [C(x_{0})+B_{\varepsilon }]\cap S_{r}\) for all \(x\in U\cap K\).
Since C is c.u.c. at \(x_{0}\), for the preceding positive numbers Îµ, l, c, and r, there exists a neighborhood U of \(x_{0}\) such that, for any \(x\in U\cap K\),
By taking intersection with the sphere surface \(S_{r}\) in both sides of the above formula, we can get
Step 4. For any \(x\in U\cap K\), we have \([C(x)+C(x_{0})] \cap B_{l} \subseteq C(x_{0})+B_{\varepsilon }\).
Applying (10), we can see that
This together with \(C_{1} \subseteq C_{2}\) implies that \(C(x)\subseteq C_{2}\). Furthermore, if we can prove that \(C_{2}\) is convex, then we have
This together with \(C_{2} \subseteq C_{3}\) implies that
On the other hand, for any \(u\in C_{3}\cap B_{l}\) with \(u\neq 0\), we have \(0<\u\<l\). By the definition of \(C_{3}\), there exists \(\lambda >0\) such that \(\lambda u\in [C(x_{0})+B_{\varepsilon }] \cap S_{l}\). Thus we can see that \(\lambda >1\). This together with \(\lambda u\in C(x_{0})+B_{\varepsilon }\) implies that \(u\in C(x_{0})+B _{\varepsilon }\). Then, by the arbitrariness of u, we get
We conclude from (11) and (12) that, for any \(x\in U\cap K\),
Finally, we show that \(C_{2}\) is convex. Since \(C_{2}\) is a cone, we only need to prove that, for any \(w_{1},w_{2}\in C_{2}\setminus \{0\}\), one has \(w_{1}+w_{2}\in C_{2}\). According to the definition of \(C_{2}\), there exist \(\lambda _{1},\lambda _{2}>0\) such that \(\lambda _{1}w_{1},\lambda _{2}w_{2}\in C(x_{0})+B_{\varepsilon }\) and \(\xi (\lambda _{1}w_{1})=\xi (\lambda _{2}w_{2})=c\). Let \(\lambda = \lambda _{1}/(\lambda _{1}+\lambda _{2})\), which implies that \(0<\lambda <1\). By the linearity of Î¾, we have
Noting that \(C(x_{0})+B_{\varepsilon }\) is convex, we can see that \((1\lambda )\lambda _{1}w_{1}+\lambda \lambda _{2}w_{2}\in C(x_{0})+B _{\varepsilon }\). This together with (13) implies that \((1\lambda )\lambda _{1}w_{1}+\lambda \lambda _{2}w_{2}\in C_{2}\). Hence,
which implies that \(w_{1}+w_{2}\in C_{2}\).â€ƒâ–¡
For the simplification of the proof for our main results in this paper, we also need the following lemma.
Lemma 3.2
Let \((t_{0},x_{0},y_{0})\in E\times K\times K\). Assume that F is lower \(\{C(x_{0})\}\)continuous at \(x_{0}\), C is c.u.c. at \(x_{0}\), and \(C(x_{0})\) has a bounded and closed base. Then, for any net \(\{(t_{\alpha },x_{\alpha },y_{\alpha })\}\to (t_{0},x_{0},y_{0})\) with \(F(t_{\alpha },x_{\alpha },y_{\alpha })\subseteq C(x_{\alpha })\), we have \(F(t_{0},x_{0},y_{0})\subseteq C(x_{0})\).
Proof
Suppose to the contrary that there exists \(w_{0}\in F(t_{0},x_{0},y _{0})\) such that \(w_{0}\notin C(x_{0})\). Since \(C(x_{0})\) is closed, there exists \(\varepsilon >0\) such that
Applying the lower \(\{C(x_{0})\}\)continuity of F, there exists \(\alpha _{1}\) such that, for all \(\alpha >\alpha _{1}\),
This implies that
Take any real number \(l>\varepsilon \) such that \(w_{0}+B_{\varepsilon }\subseteq B_{l}\). Since C is c.u.c. at \(x_{0}\in K\) and \(C(x_{0})\) has a bounded closed base, it follows from LemmaÂ 3.1 that there exists \(\alpha _{2}\) such that, for any \(\alpha >\alpha _{2}\),
Let \(\alpha _{0}>\alpha _{1}\) and \(\alpha _{0}>\alpha _{2}\). Then, for any \(\alpha >\alpha _{0}\), we conclude from (15) and \(w_{0}+B _{\varepsilon }\subseteq B_{l}\) that
Combining (16) with (17), we can see that \([w_{0}+B_{\varepsilon }]\cap [C(x_{0})+B_{\varepsilon }]\neq \emptyset \), which contradicts (14). Hence, \(F(t_{0},x_{0},y_{0}) \subseteq C(x_{0})\).â€ƒâ–¡
Remark 3.1
In LemmaÂ 3.2, we use the assumptions that C is c.u.c. and F is lower \(\{C(x)\}\)continuous, which is very different from the ones that C is closed and F is l.s.c. frequently used in the previous literature.
Now, we are ready to present our main results.
Theorem 3.1
Let \(F(t,x,x)\subseteq C(x)\) for any \((t,x)\in E\times K\). Assume that \(\varOmega =\{x\in K:x\in Q(x)\}\) is closed, \(Q^{1}(y)\) is open in K for any \(y\in K\), and the following conditions are satisfied:

(i)
C is c.u.c. on K, and for any \(x\in K\), \(C(x)\) has a bounded and closed base;

(ii)
T is l.s.c. on \(K\times K\) with nonempty and convex values;

(iii)
for any \(y\in K\), the mapping \((t,x)\to F(t,x,y)\) is lower \(\{C(x)\}\)continuous on \(E\times K\);

(iv)
for any \(x\in K\), the mapping \((t,y)\to F(t,x,y)\) is upper properly \(C(x)\)quasiconvex on \(E\times K\);

(v)
there exist a nonempty and compact subset M of K and a nonempty, compact, and convex subset N of K such that, for each \(x\in K\setminus M\), there exists some \(y\in N\cap Q(x)\) such that \(F(t,x,y)\nsubseteq C(x)\) for some \(t\in T(x,x)\).
Then \(\operatorname{Sol}(F_{1})\) is nonempty.
Proof
Define two setvalued mappings \(A,H:K\rightrightarrows K\) as follows: for each \(x\in K\),
and
We assert that there exists \(\bar{x}\in K\) such that \(H(\bar{x})= \emptyset \). If \(\bar{x}\in K\setminus \varOmega \), then, by the definition of H, we have \(H(\bar{x})=Q(\bar{x})=\emptyset \), which contradicts the fact that \(Q(x)\neq \emptyset \) for all \(x\in K\). Hence \(\bar{x}\in \varOmega \) and \(Q(\bar{x})\cap A(\bar{x})=\emptyset \). This implies that \(\bar{x}\in Q(\bar{x})\), and for any \(y\in Q(\bar{x})\) and any \(t\in T(\bar{x},\bar{x})\), we have \(F(t,\bar{x},y)\subseteq C( \bar{x})\), and so xÌ„ is a solution of (GSVQEP 1). Therefore, it only suffices to prove that there exists \(\bar{x}\in K\) such that \(H(\bar{x})=\emptyset \).
The proof is divided into three steps.
Step 1. For any \(x\in K\), \(x\notin \operatorname{conv}(H(x))\).
In fact, we can firstly show that \(A(x)\) is convex. Indeed, for any \(y_{1},y_{2}\in A(x)\), there exist \(t_{1},t_{2}\in T(x,x)\) such that
Set \(y_{\lambda }=\lambda y_{1}+(1\lambda )y_{2}\) and \(t_{\lambda }= \lambda t_{1}+(1\lambda )t_{2}\) for any \(0<\lambda <1\). Since K and \(T(x,x)\) are both convex, we have \(y_{\lambda }\in K\) and \(t_{\lambda }\in T(x,x)\). If \(F(t_{\lambda },x,y_{\lambda })\subseteq C(x)\), then, by the upper properly \(C(x)\)quasiconvexity of \(F(\cdot ,x,\cdot )\), we know that there exists \(i_{0}\in \{1,2\}\) such that
This contradicts (18). Hence \(F(t_{\lambda },x,y_{\lambda })\nsubseteq C(x)\). It follows that \(y_{\lambda }\in A(x)\). Then \(A(x)\) is convex.
Next, we show that \(x\notin \operatorname{conv}(H(x))\). In fact, for any given \(x\in K\), since \(Q(x)\) and \(A(x)\) are convex, we can easily see that \(H(x)\) is also convex. Thus, we consider the next two cases. (1)Â If \(x\in \varOmega \), then the assumption \(F(t,x,x)\subseteq C(x)\) implies \(x\notin A(x)\). Hence \(x\notin \operatorname{conv}(A(x))\). This together with \(H(x)\subseteq A(x)\) implies that \(x\notin \operatorname{conv}(H(x))\). (2)Â If \(x\in K\setminus \varOmega \), then \(x\notin Q(x)\). This together with \(\operatorname{conv}(Q(x))=\operatorname{conv}(H(x))\) implies that \(x\notin \operatorname{conv}(H(x))\).
Step 2. For all \(y\in X\), \(H^{1}(y)\) is open in K.
Indeed, for each \(y\in K\), by the definitions of A and H, we have
and
Noting that Î© is closed and \(Q^{1}(y)\) is open in K, it suffices to show that \(A^{1}(y)\) is also open in K. Thus, we only need to show that \(K\setminus A^{1}(y)\) is closed. In fact, for any net \(\{x_{\alpha }\}\subseteq K\setminus A^{1}(y)\) with \(x_{\alpha }\to x_{0}\in K\), if \(x_{0}\in A^{1}(y)\), then there exists \(t_{0}\in T(x_{0},x_{0})\) such that
We conclude from condition (ii) and LemmaÂ 2.1(ii) that there exists a net \(\{t_{\alpha }\}\) such that \(t_{\alpha }\to t_{0}\) with \(t_{\alpha }\in T(x_{\alpha },x_{\alpha })\). Since \(x_{\alpha }\notin A^{1}(y)\), we can see that
This together with \(t_{\alpha }\in T(x_{\alpha },x_{\alpha })\) implies that \(F(t_{\alpha },x_{\alpha },y)\subseteq C(x_{\alpha })\). Then, by conditions (i), (iii) and LemmaÂ 3.2, we have \(F(t_{0},x_{0},y) \subseteq C(x_{0})\), which contradicts (19). Hence \(x_{0}\notin K\setminus A^{1}(y)\).
Step 3. There exist a nonempty and compact subset M of K and a nonempty, compact, and convex subset N of K such that, for each \(x\in K\setminus M\), \(N\cap H(x)\neq \emptyset \).
Indeed, by condition (v), there exist a nonempty and compact subset M of K and a nonempty, compact, and convex subset N of K such that, for each \(x\in K\setminus M\), there exists \(y\in N\cap Q(x)\) such that \(F(t,x,y)\nsubseteq C(x)\) for some \(t\in T(x,x)\). Hence \(y\in Q(x)\cap A(x)\cap N\). It follows that \(y\in N\cap H(x)\).
Thus all the conditions of LemmaÂ 2.3 are satisfied. We conclude from LemmaÂ 2.3 that there exists \(\bar{x}\in X\) such that \(H(\bar{x})=\emptyset \).â€ƒâ–¡
In the following, the closedness of the solution set \(\operatorname{Sol}(F_{1})\) is discussed.
Lemma 3.3
Suppose that the conditions in TheoremÂ 3.1 are all satisfied. Assume that Q is l.s.c. on K and F is lower \(\{C(x)\}\)continuous on \(E\times K\times K\). Then \(\operatorname{Sol}(F_{1})\) is closed.
Proof
In fact, for any net \(\{x_{\alpha }\}\subseteq \operatorname{Sol}(F_{1})\) with \(x_{\alpha }\to x_{0}\in K\), we can see that \(x_{\alpha }\in Q(x_{ \alpha })\), which implies that \(x_{\alpha }\in \varOmega \). Since Î© is closed in K, we have \(x_{0}\in \varOmega \), i.e., \(x_{0}\in Q(x_{0})\). If \(x_{0}\notin \operatorname{Sol}(F_{1})\), then there exist some \(y_{0}\in Q(x_{0})\) and \(t_{0}\in T(x_{0},x_{0})\) such that
Since Q and T are l.s.c. on K and \(K\times K\), respectively, we conclude from LemmaÂ 2.1(ii) that there exist \(y_{\alpha } \in Q(x_{\alpha })\) and \(t_{\alpha }\in T(x_{\alpha },x_{\alpha })\) such that \(y_{\alpha }\to y_{0}\) and \(t_{\alpha }\to t_{0}\). Applying the assumption that \(x_{\alpha }\in \operatorname{Sol}(F_{1})\), we can get \(F(t,x_{ \alpha },y)\subseteq C(x_{\alpha })\) for all \(y\in Q(x_{\alpha })\) and \(t\in T(x_{\alpha },x_{\alpha })\). Therefore,
By conditions (i) and (iii), we conclude from LemmaÂ 3.2 and (21) that
which contradicts (20). Hence \(x_{0}\in \operatorname{Sol}(F_{1})\), which implies that \(\operatorname{Sol}(F_{1})\) is a closed subset of K.â€ƒâ–¡
Remark 3.2
TheoremÂ 3.1 is very different from TheoremÂ 4.2 of Lin and Huang [18]. More specifically, (a)Â in order to establish the existence result of solutions for (GSVQEP 1), Lin and Huang used the upper semicontinuity condition for the conevalued mapping, while we used here the cosmically upper continuity condition for it; (b)Â in TheoremÂ 4.2 of Lin and Huang[18], the setvalued mapping \(F(\cdot ,\cdot ,y)\) is assumed to be l.s.c. on \(E\times K\), while in TheoremÂ 3.1 of this paper, it is assumed to be lower \(\{C(x)\}\)continuous on \(E\times K\).
Remark 3.3
Chen, Yang, and Yu [26] discussed another form of equilibrium problem with variable ordering structure. By applying the scalarization method and the conditions that the conevalued mappings \(C(\cdot )\) and \(W(\cdot ):=Z\setminus \operatorname{int}C(\cdot )\) are both u.s.c., they proved an existence theorem of solutions for a singlevalued weak vector quasiequilibrium problem with variable ordering structure. Different from TheoremÂ 3.2 of Chen, Yang, and Yu [26], in TheoremÂ 3.1 of this paper, by using the condition that C is c.u.c., we also obtain the existence result of solutions for a setvalued strong vector quasiequilibrium problem with variable ordering structure.
Example 3.1
Let \(Z=\mathbb{R}^{2}\), \(X=Y=\mathbb{R}\), \(K=[0,\pi /2]\), and \(E=[0,1]\). For each \(x,y\in K\) and \(t\in E\), let
and
For any given \(y_{0}\in (0,\pi /2)\) and \(t_{0}\in E\), if \(x_{0}=y_{0}\), then \(h(x_{0},y_{0})=0\), and so
We can verify that \(F(\cdot ,\cdot ,y_{0})\) is not l.s.c. at the point \((t_{0},x_{0})\). In fact, for any neighborhood \(V\times U\) of \((t_{0},x_{0})\), there exists \((t_{0},\tilde{x})\in V\times U\) such that \(\tilde{x}>x_{0}=y_{0}\). It follows from the definition of F that
Clearly, there exist a point \(v_{0}\in F(t_{0},x_{0},y_{0})\) with \(\v_{0}\>5\) and a neighborhood W of \(v_{0}\) such that \(W\cap \operatorname{cl}(B_{5})=\emptyset \). This together with (22) implies that \(W\cap F(t_{0},\tilde{x},y_{0})=\emptyset \). It is easy to know that \(F(\cdot ,\cdot ,y_{0})\) is not l.s.c. at the point \((t_{0},x_{0})\). Therefore, the condition that F is l.s.c. on \(E\times K\times K\) is not satisfied in TheoremÂ 4.2 of [18]. So, we are unable to use TheoremÂ 4.2 of [18] to decide the existence of the solutions for (GSVQEP 1).
However, for this example, we can verify that all the conditions in TheoremÂ 3.1 of this paper are satisfied.
Indeed, for each \(x\in K\), it is easy to see that \(C(x)\subseteq \mathbb{R}^{2}\) is a closed, convex, and pointed cone and Î© is closed as \(\varOmega =[0,1]\). For each \(y\in K\), \(Q^{1}(y)\) is open in K because \(Q^{1}(y)=K\) or \(Q^{1}(y)=\emptyset \). In addition, for any \((t,x)\in E\times K=[0,1]\times [0,\pi /2]\), by the definitions of C and F, we have
In the following, we verify that conditions (i)â€“(v) of TheoremÂ 3.1 are all satisfied. By the assumptions of C and T, it is easy to get that conditions (i) and (ii) hold. Thus we consider whether the mapping \((t,y)\to F(t,x,y)\) is upper properly \(C(x)\)quasiconvex on \(E\times K\) for any \(x\in K\). In fact, let \(x\in K\), \(\lambda \in (0,1)\), \((t_{1},y_{1}),(t_{2},y_{2})\in E\times K\), and \(y_{2}\geq y_{1}\). Set \(y_{\lambda }=\lambda y_{1}+(1\lambda )y_{2}\) and \(t_{\lambda }=\lambda t_{1}+(1\lambda )t_{2}\). It suffices to show that there exists \(i_{0}\in \{1,2\}\) such that \(F(t_{i_{0}},x,y_{i _{0}})\subseteq F(t_{\lambda },x,y_{\lambda })+C(x)\). By the definition of C, we have
By the definition of F, we can see that \(h(x,y)\in C(x)\). This together with (24) implies that
If \(x> y_{1}\), from (24) and (25), we can see that
If \(x\leq y_{1}\), then \(y_{2}\geq y_{\lambda }\geq y_{1}\), and so \(h(x,y_{\lambda })h(x,y_{1})=((y_{1}y_{\lambda })\cos {x},(y_{1}y _{\lambda })\sin {x})\in C(x)\). It follows that \(h(x,y_{1})+C(x) \subseteq h(x,y_{\lambda })+C(x)\). This together with (24) implies that
Therefore, the mapping \((t,y)\to F(t,x,y)\) is upper properly \(C(x)\)quasiconvex on \(E\times K\) for any \(x\in K\).
Next, we show that \((t,x)\to F(t,x,y)\) is lower \(\{C(x)\}\)continuous on \(E\times K\) for any given \(y\in K\). Let \((t_{0},x_{0})\in E\times K\) and \(\varepsilon >0\). We consider the following two cases. a) Assume that \(x_{0} > y\). Since \(F(t,x,y)=\operatorname{cone}\{(\cos {tx},\sin {tx}) \}\cap \operatorname{cl}(B_{5})\) for any \(x> y\), it is obvious that \(F(\cdot ,\cdot ,y)\) is lower \(\{C(x_{0})\}\)continuous at \((t_{0},x_{0})\in E\times K\). b) Assume that \(x_{0} > y\). We firstly define a vectorvalued mapping \(g:K\to Z\) as follows:
Since \(x_{0}\leq y\), we conclude from the continuity of \(h(\cdot , y)\) and \(h(y,y)=(0,0)\) that g is continuous on K. Thus there exists a neighborhood W of \(x_{0}\) such that, for all \(x\in U(x_{0})\), \(g(x_{0})\in g(x)+ B_{\varepsilon }\). Hence,
This implies that \((t,x)\to F(t,x,y)\) is lower \(\{C(x)\}\)continuous on \(E\times K\).
Since K is a nonempty, compact, and convex subset of X, it is easy to see that condition (v) in TheoremÂ 3.1 is satisfied with \(M=N=K\).
Therefore, conditions (i)â€“(v) in TheoremÂ 3.1 of this paper are all satisfied. Thus, by TheoremÂ 3.1, (GSVQEP 1) has at least one solution in K.
In fact, we can also verify that \(\bar{x}=1\) is the solution of this example. For all \(t\in T(\bar{x})=[0,1]\) and for all \(y\in Q(\bar{x})=[0,1]\), we conclude from \(y\leq 1=\bar{x}\) and \(t\in [0,1]\) that
This indicates that \(\bar{x}=1\) solves (GSVQEP 1).
Next, the existence of solutions and the closedness of the solution set for (GSVQEP 2) are discussed.
Theorem 3.2
Let X, Y, Z, E, K, C, T, Q, and F be the same as in TheoremÂ 3.1. Suppose that E is compact and Î© is closed. If the following conditions are satisfied:

(i)
C is c.u.c. on K, and for all \(x\in K\), \(C(x)\) has a bounded and closed base;

(ii)
T is u.s.c. on \(K\times K\) with nonempty and closed values;

(iii)
for each \(y\in K\), the mapping \((t,x)\to F(t,x,y)\) is lower \(\{C(x)\}\)continuous on \(E\times K\);

(iv)
for any \((t,x)\in E\times K\), the mapping \(y\to F(t,x,y)\) is upper properly \(C(x)\)quasiconvex on K;

(v)
there exist a nonempty and compact subset M of K and a nonempty, compact, and convex subset N of K such that, for each \(x\in K\setminus M\), there exists some \(y\in N\cap Q(x)\) such that \(F(t,x,y)\nsubseteq C(x)\) for all \(t\in T(x,x)\).
Then \(\operatorname{Sol}(F_{2})\) is nonempty.
Proof
For each \(x\in K\), let
and
Similar to the analysis as TheoremÂ 3.1, it suffices to be proved that there exists \(\bar{x}\in K\) such that \(H(\bar{x})=\emptyset \).
The proof is divided into three steps.
Step 1. For all \(x\in K\), \(x\notin \operatorname{conv}(H(x))\).
In fact, for every \(t\in T(x,x)\), by the similar arguments for the convexity of A in the proof of TheoremÂ 3.1, we can show that \(A(x)\) is convex for each \(x\in K\). In the same way as in TheoremÂ 3.1, we can show that \(x\notin \operatorname{conv}(H(x))\).
Step 2. For all \(y\in X\), \(H^{1}(y)\) is open in K.
By the definition of A and H, we can see that \(A^{1}(y)=\{x \in K:\forall t\in T(x,x), F(t,x,y)\nsubseteq C(x)\}\) and \(H^{1}(y) = Q^{1}(y)\cap [( K\setminus \varOmega )\cup A^{1}(y)]\). Similar to the proof of openness of \(H^{1}(y)\) in TheoremÂ 3.1, it suffices to show that \(K\setminus A^{1}(y)\) is closed.
In fact, for any net \(\{x_{\alpha }\}\subseteq K\setminus A^{1}(y)\) with \(x_{\alpha }\to x_{0}\in K\), we have \(x_{\alpha }\notin A^{1}(y)\). Thus, for each Î±, there exists some \(t_{\alpha }\in T(x_{\alpha },x_{\alpha })\subseteq E\) such that
Since E is compact, we can see that T is u.s.c. with compact values. By LemmaÂ 2.1(i), there exists a subnet of \(\{t_{ \alpha }\}\) which converges to \(t_{0}\in T(x_{0},x_{0})\). Without loss of generality, we assume that \(t_{\alpha }\to t_{0}\). It follows from (26) and LemmaÂ 3.2 that \(F(t_{0},x_{0},y)\subseteq C(x _{0})\). Hence, \(x_{0}\in K\setminus A^{1}(y)\), which implies that \(K\setminus A^{1}(y)\) is closed.
Step 3. There exist a nonempty and compact subset M of K and a nonempty, convex, and compact subset N of K such that, for each \(x\in K\setminus M\), \(N\cap H(x)\neq \emptyset \).
Indeed, by condition (v), there exist a nonempty compact subset M of K and a nonempty compact convex subset N of K such that, for each \(x\in K\setminus M\), there exists some \(y\in N\cap Q(x)\) such that \(F(t,x,y)\nsubseteq C(x)\) for all \(t\in T(x,x)\). It follows that \(y\in Q(x)\cap A(x)\), and so \(y\in N\cap H(x)\).
Applying LemmaÂ 2.3, there exists \(\bar{x}\in K\) such that \(H(\bar{x})=\emptyset \).â€ƒâ–¡
Lemma 3.4
Let \(\operatorname{Sol}(F_{2})\neq \emptyset \). Suppose that Q is l.s.c. on K, T is u.s.c. on \(K\times K\) with compact values, and F is lower \(\{C(x)\}\)continuous on \(E\times K\times K\). Then \(\operatorname{Sol}(F_{2})\) is closed.
Proof
In fact, for any net \(\{x_{\alpha }\}\subseteq \operatorname{Sol}(F_{2})\) with \(x_{\alpha }\to x_{0}\in K\), we need to prove that \(x_{0}\in \operatorname{Sol}(F _{2})\). If it is not true, then there exists \(y_{0}\in Q(x_{0})\) such that, for all \(t\in T(x_{0},x_{0})\),
Since \(y_{0}\in Q(x_{0})\) and Q is l.s.c. on K, applying LemmaÂ 2.1(ii), there exists \(y_{\alpha }\in Q(x_{\alpha })\) such that \(y_{\alpha }\to y_{0}\). We conclude from \(x_{\alpha }\in \operatorname{Sol}(F_{2})\) and \(y_{\alpha }\in Q(x_{\alpha })\) that there exists \(t_{\alpha }\in T(x _{\alpha },x_{\alpha })\) such that
Noting that T is u.s.c. with compact values, by LemmaÂ 2.1(i), there exists a subnet of \(\{t_{\alpha }\}\) which converges to \(t_{0}\in T(x_{0},x_{0})\). Without loss of generality, let \(t_{\alpha }\to t_{0}\). It follows from (28) and LemmaÂ 3.2 that \(F(t_{0},x_{0},y_{0})\subseteq C(x_{0})\), which contradicts (27). Thus \(x_{0}\in \operatorname{Sol}(F_{2})\). This implies that \(\operatorname{Sol}(F_{2})\) is closed.â€ƒâ–¡
Remark 3.4
If K is a nonempty, compact, and convex subset of X, we can take \(M=N=K\). Then, condition (v) in Theorems 3.1 and 3.2 is satisfied trivially. Furthermore, if Q is u.s.c. with nonempty, closed, and convex values, then the set \(\varOmega =\{x\in K:x\in Q(x)\}\) is closed.
Remark 3.5
In the situation that \(Q(x)\equiv K\) for all \(x\in K\), Fu and Wang [29] obtained an existence theorem of solutions for (GSVQEP 2)â€”TheoremÂ 3.1, which is very different from TheoremÂ 3.2 of this paper. Above all the differences, the continuity condition on the conevalued mapping C is the biggest one. Indeed, the conevalued mapping C is assumed to be u.s.c. in TheoremÂ 3.1 of Fu and Wang [29], while it is assumed to be c.u.c. in TheoremÂ 3.2 of this paper.
By applying TheoremÂ 3.1, we can get the following corollary.
Corollary 3.1
Let X, Z, K, and C be same as in TheoremÂ 3.1. Let \(\widetilde{F}:K \times K\rightrightarrows Z\) be a setvalued mapping. Suppose that the following conditions are satisfied:

(i)
C is c.u.c. on K, and for all \(x\in K\), \(C(x)\) has a bounded and closed base;

(ii)
for all \(x\in K\), \(\widetilde{F}(x,x)\subseteq C(x)\);

(iii)
for each \(x\in K\), the mapping \(y\to \widetilde{F}(x,y)\) is upper properly \(C(x)\)quasiconvex on K;

(iv)
for each \(y\in K\), the mapping \(x\to \widetilde{F}(x,y)\) is lower \(\{C(x)\}\)continuous on K;

(v)
there exist a nonempty and compact subset M of K and a nonempty, compact, and convex subset N of K such that, for each \(x\in K\setminus M\), there exists \(y\in N\cap Q(x)\) satisfying \(\widetilde{F}(x,y)\nsubseteq C(x)\).
Then the solution set \(\operatorname{Sol}(\widetilde{F})\) is nonempty and closed.
Proof
Let \(Y=X\) and \(E=K\). For each \(x,y\in K\), let
It is easy to verify that conditions (i)â€“(v) of TheoremÂ 3.1 are all satisfied. Then, by TheoremÂ 3.1, there exists \(\bar{x} \in K\) such that
Hence, xÌ„ is a solution of (SVEP).
Next, we show that \(\operatorname{Sol}(\widetilde{F})\) is closed. In fact, for any net \(\{x_{\alpha }\}\subseteq \operatorname{Sol}(\widetilde{F})\) with \(x_{\alpha }\to x _{0}\in K\), we can see that
Then, for each \(y\in X\), we conclude from (i), (iv), and LemmaÂ 3.2 that \(\widetilde{F}(x_{0},y)\subseteq C(x_{0})\). This implies that \(x_{0}\in \operatorname{Sol}(\widetilde{F})\). Hence, \(\operatorname{Sol}(\widetilde{F})\) is closed.â€ƒâ–¡
4 Conclusions
The aim of this paper is to investigate generalized strong vector quasiequilibrium problems with variable ordering structure, which has already been studied in the previous literature. However, we would like to point out that the concept of upper semicontinuity, which is widely used for conevalued mapping in many results, just as Borde and Crouzeix [33] pointed out, is actually not suitable for the conevalued mapping because of its strange characterization. And so, the value of those previous results will be heavily discounted. In order to deal with conevalued mapping, some papers consider the existence of solutions for equilibrium problems under the conditions that conevalued mapping is closed and the equilibrium mapping is continuous vectorvalued mapping or lower semicontinuous setvalued mapping. Different from these existing literature sources, we establish the existence of solutions for (GSVQEP) by using the assumption that the equilibrium mapping is conecontinuous. Another difficulty of this article is the continuity condition of conevalued mapping. Without using the upper semicontinuity, we consider the inclusion relationship of bases for these cones and further obtain a local property by using the modified concept called cosmically upper continuity. Also, this conclusion plays a key role in the existence theorems of solution for these (GSVQEP). Moreover, an example is given to illustrate the validity of our theorems. These results obtained in this paper extend and develop some recent works in this field.
References
Blum, E., Oettli, W.: From optimization and variational inequalities to equilibrium problems. Math. Stud. 63, 123â€“145 (1994)
FloresBazÃ¡n, F.: Existence theorems for generalized noncoercive equilibrium problems: quasiconvex case. SIAM J. Optim. 11, 675â€“790 (2000)
FloresBazÃ¡n, F., FloresBazÃ¡n, F.: Vector equilibrium problems under asymptotic analysis. J. Glob. Optim. 26, 141â€“166 (2003)
Anh, L.Q., Bantaojai, T., Duc, N.P., Duy, T.Q., Wangkeeree, R.: Convergence of solutions to lexicographic equilibrium problems. Appl. Numer. Optim. 1, 39â€“51 (2019)
Ceng, L.C.: Convergence analysis of a Mannlike iterative algorithm in reflexive Banach spaces. Appl. SetValued Anal. Optim. 1, 1â€“18 (2019)
Saidi, A., Chadli, O., Yao, J.C.: Second order nonlinear evolution equations with time dependent pseudomonotone and quasimonotone operators: an equilibrium problem approach. Appl. Anal. Optim. 1, 245â€“259 (2017)
Bianchi, M., Pini, R.: Coercivity conditions for equilibrium problems. J. Optim. Theory Appl. 124, 79â€“92 (2005)
Iusem, A.N., Sosa, W.: New existence results for equilibrium problems. Nonlinear Anal. TMA 52, 621â€“635 (2003)
Iusem, A.N., Kassay, G., Sosa, W.: On certain conditions for the existence of solutions of equilibrium problems. Math. Program. 116, 259â€“273 (2009)
GutiÃ©rrez, C., Kassay, G., Novo, V., RÃ³denasPedregosa, J.L.: Ekeland variational principles in vector equilibrium problems. SIAM J. Optim. 27, 2405â€“2425 (2017)
Ansari, Q.H.: Vector equilibrium problems and vector variational inequalities. In: Giannessi, F. (ed.) Vector Variational Inequalities and Vector Equilibria, pp.Â 1â€“16. Kluwer Academic, Dordrecht (2000)
Ansari, Q.H., Farajzadeh, A.P., Schaible, S.: Existence of solutions of strong vector equilibrium problems. Taiwan. J. Math. 16, 165â€“178 (2012)
Chen, J., Liou, Y.C., Wan, Z., Yao, J.C.: A proximal point method for a class of monotone equilibrium problems with linear constraints. Oper. Res. 15, 275â€“288 (2015)
Kassay, G.: On equilibrium problems. In: Chinchuluun, A., Pardalos, P., Enkhbat, R., Tseveendorj, I. (eds.) Optimization and Optimal Control: Theory and Applications, vol.Â 39, pp.Â 55â€“83. Springer, Berlin (2010)
Guu, S.M., Li, J.: Vector quasiequilibrium problems: separation, saddle points and error bounds for the solution set. J. Glob. Optim. 58, 751â€“767 (2014)
Castellani, M., Giuli, M.: An existence result for quasiequilibrium problems in separable Banach spaces. J. Math. Anal. Appl. 425, 85â€“95 (2015)
Alleche, B., RÇŽdulescu, V.D.: Solutions and approximate solutions of quasiequilibrium problems in Banach spaces. J. Optim. Theory Appl. 170, 629â€“649 (2016)
Lin, L.J., Huang, Y.J.: Generalized vector quasiequilibrium problems with applications to common fixed point theorems and optimization problems. Nonlinear Anal. TMA 66, 1275â€“1289 (2007)
Zhang, W.B., Shan, S.Q., Huang, N.J.: Existence of solutions for generalized vector quasiequilibrium problems in abstract convex spaces with applications. Fixed Point Theory Appl. 2015, 29 (2015)
Ansari, Q.H., FloresBazÃ¡n, F.: Generalized vector quasiequilibrium problems with applications. J. Math. Anal. Appl. 277, 246â€“356 (2003)
Hai, N.X., Khanh, P.Q.: Existence of solutions to general quasiequilibrium problems and applications. J. Optim. Theory Appl. 133, 317â€“327 (2007)
Ansari, Q.H., Yao, J.C.: Systems of vector quasiequilibrium problems and their applications. In: Burachik, R.S., Yao, J.C. (eds.) Variational Analysis and Generalized Differentiation in Optimization and Control, vol.Â 47, pp.Â 1â€“42. Springer, New York (2010)
Luo, Z.Q., Pang, J.S., Ralph, D.: Mathematical Programs with Equilibrium Constraints. Cambridge University Press, Cambridge (1996)
Anh, L.Q., Hung, N.V.: Levitinâ€“Polyak wellposedness for strong bilevel vector equilibrium problems and applications to traffic network problems with equilibrium constraints. Positivity 22, 1223â€“1239 (2018)
Yu, P.L.: Cone convexity, cone extreme points, and nondominated solutions in decision problems with multiobjectives. J. Optim. Theory Appl. 14, 319â€“377 (1974)
Chen, G.Y., Huang, X.X., Yang, X.Q.: Vector Optimization, SetValued and Variational Analysis. Springer, Berlin (2005)
Li, S.J., Teo, K.L., Yang, X.Q., Wu, S.Y.: Gap functions and existence of solutions to generalized vector quasiequilibrium problems. J. Glob. Optim. 34, 427â€“440 (2006)
Mastroeni, G.: On the image space analysis for vector quasiequilibrium problems with a variable ordering relation. J. Glob. Optim. 53, 203â€“214 (2012)
Fu, J.Y., Wang, S.H.: Generalized strong vector quasiequilibrium problem with domination structure. J. Glob. Optim. 55, 839â€“847 (2013)
Farajzadeh, A., Lee, B.S., Plubteing, S.: On generalized quasivector equilibrium problems via scalarization method. J. Optim. Theory Appl. 168, 584â€“599 (2016)
Eichfelder, G.: Optimal elements in vector optimization with a variable ordering structure. J. Optim. Theory Appl. 151, 217â€“240 (2011)
Anh, L.Q., Hien, D.V.: On wellposedness for parametric vector quasiequilibrium problems with moving cones. Appl. Math. 61, 651â€“668 (2016)
Borde, J., Crouzeix, J.P.: Continuity properties of the normal cone to the level sets of a quasiconvex function. J. Optim. Theory Appl. 66, 415â€“429 (1990)
Eichfelder, G.: Variable Ordering Structures in Vector Optimization. Springer, Berlin (2014)
Luc, D.T., Penot, J.P.: Convergence of asymptotic directions. Trans. Am. Math. Soc. 353, 4095â€“4121 (2001)
Aubin, J.P., Ekeland, I.: Applied Nonlinear Analysis. Wiley, New York (1984)
GÃ¶pfert, A., Riahi, H., Tammer, C., ZÄƒlinescu, C.: Variational Methods in Partially Ordered Spaces. Springer, New York (2003)
Tan, N.X.: On the existence of solutions of quasivariational inclusion problems. J. Optim. Theory Appl. 123, 619â€“638 (2004)
John, J.: Mathematical Vector Optimization in Partially Ordered Linear Spaces. Peter Lang, Frankfurt am Main (1986)
Deguire, P., Tan, K.K., Yuan, X.Z.: The study of maximal elements, fixed points for \(\mathrm{L}_{\mathrm{S}}\)majorized mappings and their applications to minimax and variational inequalities in product topological spaces. Nonlinear Anal. TMA 37, 933â€“951 (1999)
Acknowledgements
Not applicable.
Availability of data and materials
Not applicable.
Funding
This work was supported by the National Natural Science Foundation of China (Grant Nos.Â 11661055, 11201216, 11661056, 11771198, and 11801257), the Natural Science Foundation of Jiangxi Province (Grant No.Â 20151BAB201020), the China Postdoctoral Science Foundation (Grant No.Â 2015M582047), the Postdoctoral Science Foundation of Jiangxi Province (Grant No.Â 2016KY58), and the Science Fund of the Education Department of Jiangxi Province (Grant no.Â GJJ160063).
Author information
Authors and Affiliations
Contributions
MJY and WSH carried out the design of the study and performed the analysis. HJX participated in its design and coordination. All authors read and approved the final manuscript.
Corresponding author
Ethics declarations
Competing interests
The authors declare that they have no competing interests.
Additional information
Publisherâ€™s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.
About this article
Cite this article
Mao, JY., Wang, Sh. & Huang, Jx. Generalized strong vector quasiequilibrium problems with variable ordering structure. J Inequal Appl 2019, 116 (2019). https://doi.org/10.1186/s1366001920631
Received:
Accepted:
Published:
DOI: https://doi.org/10.1186/s1366001920631