On The Constrained Equilibrium Problems With "Nite Families of Players

Nonlinear Analysis 54 (2003) 525543
www.elsevier.com/locate/na
On the constrained equilibrium problems with
nite families of players
Lai-Jiu Lin
a;
, Shih Feng Cheng
a
, Xu Yao Liu
a
, Q.H. Ansari
b
a
Department of Mathematics, National Changhua University of Education,
Changhua, Taiwan 50058, Peoples Republic of China
b
Department of Mathematics, Aligarh Muslim University, Aligarh 202 002, India
Received 30 January 2002; accepted 22 November 2002
Abstract
In this paper, we consider the equilibrium problem with nite number of families of players
such that each family may not have the same number of players and nite number of families
of constrained correspondences on the strategy sets. We also consider the case with two nite
families of constrained correspondences on the strategies sets. We demonstrate an example of
our equilibrium problem. We derive a xed point theorem for a family of multimaps and a
coincidence theorem for two families of multimaps. By using these results, we establish the
existence of a solution of our equilibrium problems. The results of this paper generalize some
known results in the literature.
? 2003 Elsevier Science Ltd. All rights reserved.
Keywords: Constrained equilibrium problems; Debreu Social equilibrium problem; Nash equilibrium problem;
Fixed point; Coincidence theorem
1. Introduction
Let I be any nite index set and for each k I , let J
k
be a nite index set. We
consider the multiobjective game in the strategy form = (X
k
j
; A
k
j
; F
k
j
)
kI; jJ
k
with
nite families of players such that each family may not have the same (nite) number
of players and each player in any family have multivalued gain (or payo) function.
In this research, the rst three authors were supported by the National Science Council of the Republic
of China.
Corresponding author. Tel.: +886-4-7232105 ext. 3219; fax: +886-4-7211192.

E-mail address: maljlin@math.ncue.edu.tw (L.-J. Lin).
0362-546X/03/$ - see front matter ? 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0362-546X(03)00110-X
526 L.-J. Lin et al. / Nonlinear Analysis 54 (2003) 525543
For each k I and j J
k
, let X
k
j
denote the strategy set of jth player in kth family,
Y
k
=
jJ
k
X
k
j
, Y
k
=
lI; l=k
Y
l
, A
k
j
: Y
k
(X
k
j
the constrained correspondence which
restricts the strategy of the jth player in the kth family to the subset A
k
j
(Y
k
) of X
k
j
when all the players in other families have chosen their strategies x
ij
, i I , i = k
and j J
i
. Let F
k
j
: X
k
j
Y
k
( R
k
j
be the cost function of the jth player in the kth
family, where
k
j
is a natural number and R
k
j
is a
k
j
-dimensional Euclidean space. In
this game, the cost function F
k
j
is a multivalued function of its strategy and strategies
of players in other families. In this problem, we are interested in nding a strategies
combination y = ( y
k
)
kI

kI
Y
k
= Y, where y
k
= ( x
k
j
)
jJ
k
, with x
k
j
A
k
j
( y
k
) and
z
k
j
F
k
j
( x
k
j
; y
k
) such that
z
k
j
z
k
j
int R
k
j
+
; (1.1)
for all z
k
j
F
k
j
(x
k
j
; y
k
), x
k
j
A
k
j
( y
k
) and for each k I and j J
k
.
If we let z
k
j
= (u
1
k
j
; : : : ; u
k
j
k
j
), z
k
j
= ( u
1
k
j
; : : : ; u
k
j
k
j
), then inequality (1.1) implies that
there exists 1 6s 6
k
j
such that u
s
k
j
u
s
k
j
, that is, we can choose a collection of
objects z
k
j
from F
k
j
( x
k
j
; y
k
) such that for any collection z
k
j
chosen from F
k
j
(x
k
j
; y
k
)
there is 1 6s 6
k
j
such that u
s
k
j
u
s
k
j
.
For each k I , let J
k
= {k} be a singleton set and F
k
be a single real valued
function. Then above problem is reduced to the Debreu social equilibrium problem
[5], which is an extension of the Nash equilibrium problem [14], see also [2]. In 1952,
Debreu established the existence of a solution of social equilibrium problem. Since
then, many generalizations and applications of these two problems have been appeared
in the literature, see [2,1621] and references therein. We note that these problems
have only one family of players and the cost functions are single-valued. There is no
constrained correspondences on the strategy sets in the Nash equilibrium problem, while
Debreu social equilibrium problem has only one system of constrained correspondence
on the strategy sets. If for each k I , J
k
is not a singleton set, our problem is dierent
from Debreu social equilibrium problem. Recently, Lin [8] considered the constrained
two families of players competitive equilibrium problems and proved the existence of a
solution of this problem. In many problems, we always have nite families of players
such that each family have nite players and the cost functions are multimaps; See the
following example.
We now demonstrate an example of this kind of equilibrium problem. Let I =
{1; 2; : : : ; m} denote the index set of the companies. For each k I , let J
k
={1; 2; : : : ; n
k
}
denote the index set of factories in the kth company, F
k
j
denote the cost function of
the jth factory in the kth company. We assume that the products between the factories
in the same company are dierent, while some collections of products are the same
and some collections of products are dierent between dierent factories in dierent
companies. Therefore, the strategy of jth factory in the kth company depends on the
strategies of all factories in dierent companies. The cost function F
k
j
of the jth factory
in the kth company depends on its strategy and the strategies of factories in other
companies. With this strategies combination, each factory can choose a collection of
products which minimize the loss of each factory.
L.-J. Lin et al. / Nonlinear Analysis 54 (2003) 525543 527
In Section 2, we present some known denitions and results which will be used in the
sequel, while in Section 3, we derive a xed point theorem for a family of multimaps
and a coincidence theorem for two families of multimaps, which are used in Sections 4
and 5. In Section 4, we establish existence of a solution of our equilibrium problem
with nite number of families of players and nite number of families of constraints on
strategy sets. We note that the Nash equilibrium theorem [14] is a special case of our
result. The results of this section also generalize a result of Lin [8] which is derived by
using a xed point theorem of Park [15] and a coincidence theorem. The last section
deals with the equilibrium problem with nite number of families of players and two
nite number of families of constraints on strategy sets. In Section 4, we consider
the problem that the strategy of each family is inuenced by the strategies of other
families, while in Section 5, we consider the problem that for each k I , the strategy
of kth family is inuenced by the strategies of all the other families and the strategy
of kth family will inuence all the other families. The result of this paper generalize
some known results in the literature.
2. Preliminaries
Let X and Y be nonempty sets. A multimap T : X (Y is a function from X to the
power set of Y. Let A X, B Y, x X and y Y, we dene T(A)=
{T(x): x A};
x T
1
(y) if and only if y T(x) and T
1
(B) ={x X: T(x) B = }.
For topological spaces X and Y and A X, we denote by int A and

A the interior
and the closure of A in X, respectively. Let T : X (Y, then

T : X (Y is dened by
T = T(x) for all x X. T is said to be (i) upper semicontinuous (u.s.c.) if for every
x X and every open set U in Y with T(x) U, there exists an open neighborhood
U(x) of x such that T(x
) U for all x
U(x); (ii) lower semicontinuous (l.s.c.) if

for every x X and for every open set V in Y with T(x) V = , there exists an open
neighborhood V(x) of x such that T(x
) V = for all x
V(x); (iii) continuous if

it is both u.s.c. and l.s.c.; (iv) closed if its graph G
r
(T) = {(x; y): x X; y T(x)}
is closed in X Y; and (v) compact if there is a compact subset K Y such that
T(X) K.
A topological space is said to be acyclic if all of its reduced

Cech homology groups
vanish. A multimap T : X (Y is said to be acyclic if it is u.s.c. with acyclic compact
values. Let V(X; Y) be the family of acyclic multimaps T : X ( Y. Throughout this
paper, all topological spaces are assumed to be Hausdor.
Denition 2.1 (Luc [11]): Let Y be a topological vector space with a pointed closed
convex solid cone C, i.e., int C = , then a function : Y R is said to be mono-
tonically increasing (respectively, strict monotonically increasing) with respect to C
if (a) (b) for all a b C (respectively, (a) (b) for all a b int C).
Denition 2.2 (Luc and Vargas [12]): Let K be a nonempty convex subset of a topo-
logical vector space E and let Z be a topological vector space with a convex cone C.
A multimap G : K ( Z is said to be C-quasiconvex (respectively, C-quasiconcave) if
for any z Z, the set
{x K: there is a y G(x) such that z y C}
(respectively; {x K: there is a y G(x) such that z y C})
is convex.
G is said to be C-convex if for any x
1
; x
2
K and 0 6 61 G(x
1
)+(1)G(x
2
)
G(x
1
+ (1 )x
2
) + C.
We now mention some lemmas which will be used in the sequel.
Lemma 2.1 (Ben-El-Mechaiekh et al. [4] and Lin and Park [9]). Let X be a compact
topological space, Y a convex space, G; T : X ( Y two multimaps such that for any
x X, G(x) = and co G(x) T(x), and X =
yY
int G
1
(y), where co G(x) denotes
the convex hull of G(x). Then T has a continuous selection f: X Y, that is, there
exists a continuous function f: X Y such that f(x) T(x) for all x X.
Lemma 2.2 (Aubin and Cellina [3]). Let X and Y be two topological spaces and
T : X (Y be a multimap.
(a) If X is compact and T is u.s.c. with nonempty compact values, then T(X) is
compact.
(b) If Y is compact and T is closed, then T is u.s.c.
(c) If T is u.s.c. with nonempty closed values, then T is closed.
Lemma 2.3 (Lin and Yu [10]). Let X and Y be topological spaces, F : X Y ( R,
S : X (Y, m(x) = sup F(x; S(x)) and M(x) ={y S(x): m(x) F(x; y)}.
(a) If both F and S are l.s.c., then m is l.s.c..
(b) If both F and S are u.s.c. with nonempty compact values, then m is u.s.c..
(c) If both F and S are continuous multimaps with nonempty compact values, then
m is a continuous function and M is an u.s.c. and closed multimap.
Lemma 2.4 (Tan et al. [16]). Let X be a topological space and T : X ( R be a
continuous multimap with nonempty compact values. Then the function m: X ( R
dened by m(x) = max T(x) is continuous.
Lemma 2.5 (Park [15]). Let X be a nonempty compact and convex subset of a locally
convex topological vector space E and let T : X (X be an u.s.c. with nonempty closed
acyclic values. Then T has at least one xed point, i.e., there exists x X such that
x T( x).
Lemma 2.6 (Lee et al. [7]). Let K be a convex subset of a topological vector space
E and let D be a closed convex cone of a topological vector space Z such that
int D = . Let T : K ( Z be a multimap and for each xed e int D and any xed
a Z, let
p(z) = min{t R: z a + te D}:
If T is D-quasiconvex, then pT : X (R is R
+
-quasiconvex.
Lemma 2.7 (Fan [6]). Let X be a topological vector space and {Y
be a family
of compact spaces and Y =
. If for each , f
: X ( Y
is an u.s.c.
multimap with nonempty closed values. Then the multimap f: X ( Y dened by
f(x) =
(x) is u.s.c.
3. Fixed point and coincidence theorems
Let be any index set and for each , let E
be a topological vector space. Let

{X
be a family of nonempty convex subsets with each X
in E
, E =
,
X =
, X
; =
X
and we write X = X
. For each x X, x
denotes the th coordinate and x
the projection of x onto X
and we also write

x = (x
; x
).
Very recently, Ansari et al. [1] established the following xed point theorem for a
family of multimaps.
Theorem 3.1 (Ansari et al. [1]). For each , let
: X
( X
be a multimap.
Assume that the following conditions hold:
(a) For each and for all x
,

(x
) is a nonempty convex set;

(b) For each , X
{int
X

1
(x
): x
};
(c) There exist a nonempty compact subset K
of X
and a nonempty compact

convex subset D
of X
such that for each x
\K
there exists y
such
that x
int
X

1
(y
).
Then there exists x = ( x
X such that x

( x
) for all .
Now we establish a coincidence theorem for two families of multimaps which will
be used in Section 5 to prove the existence of a solution of equilibrium problems with
nite families of players and two nite number of families of constraints on strategy
sets.
Theorem 3.2. For each , let X
be a nonempty compact convex subset of a

locally convex topological vector space E
and let

: X
(X
and
: X
(X
be
multimaps. For each , assume that the following conditions hold:
(a) For all x
,

(x
) is a nonempty convex subset of X
;
(b) X
{int
X

1
(x
): x
};
(c) For all x
(x
) is a nonempty convex subset of X
;
(d) X
{int
X
(x
): x
}.
Then there exist x = ( x
X and u = ( u
X such that x

( u

) and
u

( x
) for all .
Proof. From conditions (a), (b) and Lemma 2.1, for each ,

: X
( X
has a
continuous selection f
: X
. Similarly, it follows from conditions (c), (d) and

Lemma 2.1 that for each ,
: X
(X
has a continuous selection g
: X
.
Let h : X X be dened by h(x) =
(g
(x
)) for x = (x
X. Then h is
a continuous function on X. Since X is a compact convex subset of a locally convex
topological vector space E =
, by Tychono xed point theorem, there exists

x = ( x
X such that x = h( x) =
(g
( x
)), that is, for each , x
=
f
(g
( x
)). For each , let u

=g
( x
), then u

X
. Hence x
=f
( u

)

( u

)
and u

= g
( x
( x
) for all .
Let I be any nite index set and for each k I , let J
k
be a nite index set. For each
k I and j J
k
, let X
k
j
be a nonempty compact convex subset of a topological vector
space E
k
j
. We write Y
k
=
jJ
k
X
k
j
, Y =
kI
Y
k
, Y
k
=
lI; l=k
Y
l
and Y =Y
k
Y
k
.
By using Theorem 3.1, we derive the following result which will be used in the
sequel.
Theorem 3.3. For each k I and j J
k
, let
k
j
: Y
k
( X
k
j
be a multimap. For each
k I , assume that the following conditions hold:
(a) For each j J
k
and for all y
k
Y
k
,
k
j
(y
k
) is a convex set;
(b) Y
k
=
int
Y
k {
jJ
k

1
k
j
(x
k
j
): (x
k
j
)
jJ
k
Y
k
}.
Then there exists y = ( y
k
)
kI
Y such that x
k
j

k
j
( y
k
) for all k I and j J
k
,
where y
k
= ( x
k
j
)
jJ
k
and y
k
= ( y
l
)
lI; l=k
.
Proof. For each k I , let g
k
: Y
k
( Y
k
be dened by g
k
(y
k
) =
jJ
k

k
j
(y
k
). Then
g
k
: Y
k
(Y
k
is a multimap with convex values.
Note that for each k I , u
k
= (v
k
j
)
jJ
k
Y
k
,
y
k
g
1
k
(u
k
) u
k
g
k
(y
k
) =
jJ
k
k
j
(y
k
)
v
k
j

k
j
(y
k
) for all j J
k
y
k
1
k
j
(v
k
j
) for all j J
k
y
k
jJ
k
1
k
j
(v
k
j
):
Therefore, g
1
k
(u
k
) =
jJ
k

1
k
j
(v
k
j
). By condition (b), for each k I ,
Y
k
=
_
{int
Y
k g
1
k
(u
k
): u
k
Y
k
}:
Therefore, for each y
k
Y
k
, there exists u
k
Y
k
such that
y
k
int
Y
k g
1
k
(u
k
) g
1
k
(u
k
)
and so u
k
g
k
(y
k
). Hence for any k I , g
k
(y
k
) is nonempty for each y
k
Y
k
. Then
from Theorem 3.1, there exists y=( y
k
)
kI
Y such that y
k
g
k
( y
k
)=
jJ
k

k
j
( y
k
).
Let y
k
= ( x
k
j
)
jJ
k
, then x
k
j

k
j
( y
k
) for all k I and j J
k
.
4. Equilibrium problems with nite families of players and nite families of
constraints on strategy sets
In this section, we establish the existence of a solution of our equilibrium problem
with nite number of families of players and nite number of families of constraints
on strategy sets by using Theorems 3.1 and 3.3.
Let I be any index set and for each k I , let J
k
be a nite index set. For each k I
and j J
k
, let X
k
j
be a nonempty compact convex subset of a topological vector space
E
k
j
. We write Y
k
=
jJ
k
X
k
j
, Y =
kI
Y
k
, Y
k
=
lI; l=k
Y
l
and Y = Y
k
Y
k
. For
each k I and j J
k
, let W
k
j
R
k
j
+
\{0} be a xed vector and W
k
=
jI
k
W
k
j
and
let F
k
j
: X
k
j
Y
k
( R
k
j
be the payo multimap and A
k
j
: Y
k
( X
k
j
be the constraints.
For each k I , let A
k
: Y
k
( Y
k
be dened by A
k
(y
k
) =
jJ
k
A
k
j
(y
k
), where y
k
=
(x
k
j
)
jJ
k
Y
k
and y
k
= (y
l
)
lI; l=k
Y
k
, also let S
W
k
: Y
k
Y
k
(R be dened by
S
W
k
(y
k
; y
k
) =
jJ
k
W
k
j
F
k
j
(x
k
j
; y
k
)
=
_
_
_
u: u =
jJ
k
W
k
j
z
k
j
; for z
k
j
F
k
j
(x
k
j
; y
k
)
_
_
_
:
For each k I , let M
W
k
: Y
k
(Y
k
be dened by
M
W
k
(y
k
) ={y
k
A
k
(y
k
): inf S
W
k
(y
k
; y
k
) = inf S
W
k
(A
k
(y
k
); y
k
)}
and
M(y) =
kI
M
W
k
(y
k
):
Throughout the paper, we shall use the above-mentioned notations, unless otherwise
specied.
k
, let F
k
j
: X
k
j
Y
k
( R
k
j
be a nonempty
multimap, A
k
j
: Y
k
( X
k
j
be a multimap with nonempty convex values such that for
all x
k
j
X
k
j
, A
1
k
j
(x
k
j
) is open in Y
k
and

A
k
j
is an u.s.c. multimap. Assume that
(i) for each k I , S
W
k
: Y
k
Y
k
( R is a continuous multimap with nonempty
compact values,
(ii) for each y
k
Y
k
, the multimap u
k
(S
W
k
(u
k
; y
k
) is R
+
-quasiconvex.
k
)
kI
Y, where y
k
= ( x
k
j
)
jJ
k
, with x
k
j

A
k
j
( y
k
) and there
exists z
k
j
F
k
j
( x
k
j
; y
k
) such that
z
k
j
z
k
j
int R
k
j
+
;
for all z
k
j
F
k
j
(x
k
j
; y
k
), x
k
j

A
k
j
( y
k
k
.
Proof. From the proof of Theorem 3.3, for each k I and y
k
Y
k
, we have A
1
k
(y
k
)=
jJ
k
A
1
k
j
(x
k
j
), where y
k
=(x
k
j
)
jJ
k
. Since for each k I and j J
k
, A
1
k
j
(x
k
j
) is open,
A
1
k
(y
k
) is open and since for each k I ,

A
k
( y
k
) =
jJ
k
A
k
j
( y
k
), by Lemma 2.7,

A
k
is an u.s.c. multimap with nonempty convex values. For each k I and for all n N,
we dene H
k; n
: Y
k
(Y
k
by
H
k; n
(y
k
) =
_
u
k
A
k
(y
k
): min S
W
k
(u
k
; y
k
) min S
W
k
(

A
k
(y
k
); y
k
) +
1
n
_
:
Since

A
k
is an u.s.c multimap with nonempty closed values, it follows from Lemma
2.2 (c) that

A
k
is closed. Since

A
k
is a multimap with nonempty compact values and
S
W
k
is a continuous multimap with compact values, it follows from Lemma 2.2 (a),
S
W
k
(

A
k
(y
k
); y
k
) is a compact set. Therefore, there exist u
k

A
k
(y
k
) and v
k
S
W
k
(u
k
; y
k
)
such that
v
k
= min S
W
k
(u
k
; y
k
) = min S
W
k
(

A
k
(y
k
); y
k
):
Since u
k

A
k
(y
k
), there exists a net u
k
A
k
(y
k
) such that u
k
u
k
. By Lemma 2.4,
the function x min S
W
k
(x; y
k
) is continuous. Therefore, for each n N
min S
W
k
(u
k
; y
k
) min S
W
k
(u
k
; y
k
)
1
n
for suciently large . This shows that
min S
W
k
(u
k
; y
k
) min S
W
k
(u
k
; y
k
) +
1
n
= min S
W
k
(

A
k
(y
k
); y
k
) +
1
n
and u
k
H
k; n
(y
k
) for suciently large . Hence for all n N and y
k
Y
k
, H
k; n
(y
k
)
is nonempty.
Since for each y
k
Y
k
, the multimap u
k
( S
W
k
(u
k
; y
k
) is R
+
-quasiconvex and
A
k
j
(y
k
) is convex, H
k; n
(y
k
) is convex.
Indeed, let u
k
; u
k
H
k; n
(y
k
) and [0; 1], then u
k
; u
k
A
k
(y
k
) and
min S
W
k
(u
k
; y
k
) min S
W
k
(

A
k
(y
k
); y
k
) +
1
n
and
min S
W
k
(u
k
; y
k
) min S
W
k
(

A
k
(y
k
); y
k
) +
1
n
:
Let m
k
and m
k
be such that
m
k
= min S
W
k
(u
k
; y
k
) S
W
k
(u
k
; y
k
)
and
m
k
= min S
W
k
(u
k
; y
k
) S
W
k
(u
k
; y
k
);
then
m
k
a and m
k
a; where a = min S
W
k
(

A
k
(y
k
); y
k
) +
1
n
:
By the denition of R
+
-quasiconvexity of the multimap u
k
( S
W
k
(u
k
; y
k
), there
exists b S
W
k
(u
k
+ (1 )u
k
; y
k
) such that
min S
W
k
(u
k
+ (1 )u
k
; y
k
) 6b min S
W
k
(

A
k
(y
k
); y
k
) +
1
n
= a:
Since for all k I and j J
k
, A
k
j
has convex values and so A
k
and therefore u
k
+
(1 )u
k
A
k
(y
k
). This shows that u
k
+ (1 )u
k
H
k; n
(y
k
) and hence H
k; n
(y
k
) is
convex.
Furthermore, for each k I and for all n N,
H
1
k; n
(u
k
) = A
1
k
(u
k
)
_
y
k
Y
k
: min S
W
k
(u
k
; y
k
) min S
W
k
(

A
k
(y
k
); y
k
) +
1
n
_
:
It is easy to see from the denition of l.s.c. that a multimap T : M ( N is l.s.c. if
and only if T
1
(V) is open in M for any open set V of N, where M and N are
topological spaces. Since for all k I , A
1
k
(y
k
) is open for all y
k
Y
k
, A
1
k
(W) =
{A
1
k
(y
k
): y
k
W} is open in Y
k
for any open set W in Y
k
, A
k
is l.s.c. Then it
is easy to see that

A
k
is also l.s.c. Since

A
k
is an u.s.c. multimap,

A
k
: Y
k
( Y
k
is a
continuous multimap with compact values. By Lemmas 2.3 and 2.4, we see that for
each y
k
Y
k
, u
k
Y
k
, the functions
y
k
min S
W
k
(

A
k
(y
k
); y
k
) and (u
k
; y
k
) min S
W
k
(u
k
; y
k
)
are continuous. Therefore, for each k I and for all n N, H
1
k; n
(u
k
) is open for all
u
k
Y
k
. Therefore,
Y
k
=
_
{H
1
k; n
(u
k
): u
k
Y
k
}
=
_
{int
Y
k H
1
k; n
(u
k
): u
k
Y
k
}:
By Theorem 3.1, there exists
n
y = (
n
y
k
)
kI
Y, where
n
y
k
= (
n
x
k
j
)
jJ
k
, with
n
x
k
j

A
k
j
(
n
y
k
) such that
min S
W
k
(
n
y
k
;
n
y
k
) min S
W
k
(A
k
(
n
y
k
);
n
y
k
) +
1
n
;
where
n
y
k
=(
n
y
l
)
lI; l=k
. Let
n
a
k
S
W
k
(
n
y
k
;
n
y
k
) be such that
n
a
k
=min S
W
k
(
n
y
k
;
n
y
k
).
Since for each k I and j J
k
, X
k
j
is compact and S
W
k
and

A
k
are u.s.c. multimaps
with compact values, it follows from Lemma 2.2 that

A
k
(Y
k
) and S
W
k
(

A
k
(Y
k
); Y
k
)
are compact. Since for each n N,
n
y
k

A
k
(Y
k
) and
n
a
k
S
W
k
(

A
k
(Y
k
); Y
k
), there ex-
ist a subnet {
n()
y
k
} of {
n
y
k
}, a subsequence {
n()
a
k
} of {
n
a
k
} and y
k

A
k
(Y
k
),
a
k
S
W
k
(

A
k
(Y
k
); Y
k
) such that
n()
( y
k
) y
k
;
n()
( y
k
) y
k
;
n()
a
k
a
k
:
As we have seen that

A
k
is a continuous multimap with nonempty compact values, by
Lemma 2.3(c), we have the function
u
k
min S
W
k
(

A
k
(u
k
); u
k
)
is continuous. Since
n()
a
k
min S
W
k
(

A
k
(
n()
y
k
);
n()
y
k
) +
1
n()
, we have
a
k
6min S
W
k
(

A
k
( y
k
); y
k
) for all k I (4.1)
by letting n() . Since for all k I ,

A
k
and S
W
k
are u.s.c. with compact values,
it follows from Lemma 2.2(c) that

A
k
and S
W
k
are closed and thus y
k

A
k
j
( y
k
) and
a
k
S
W
k
(

A
k
( y
k
); y
k
).
Inequality (4.1) shows that for each k I and j J
k
, there exist x
k
j

A
k
j
( y
k
),
z
k
j
F
k
j
( x
k
j
; y
k
) such that
a
k
=
jJ
k
W
k
j
z
k
j
6
jJ
k
W
k
j
z
k
j
for all z
k
j
F
k
j
(x
k
j
; y
k
); x
k
j

A
k
j
( y
k
): (4.2)
For each k I and j J
k
and for all x
k
j

A
k
j
( y
k
) and z
k
j
F
k
j
(x
k
j
; y
k
), let
x
ks
= x
ks
and z
ks
= z
ks
for s J
k
and s = j
in (4.2), we obtain
W
k
j
z
k
j
6W
k
j
z
k
j
;
for all z
k
j
F
k
j
(x
k
j
; y
k
), x
k
j

A
k
j
( y
k
k
.
Since W
k
j
R
k
j
+
\{0}, W
k
j
z 0 for all z 0. We have
z
k
j
z
k
j
int R
k
j
+
;
for all z
k
j
F
k
j
(x
k
j
; y
k
), x
k
j

A
k
j
( y
k
k
.
Remark 4.1. If I ={1; 2}, then Theorem 4.1 is reduced to Theorem 3.7 in [8].
k
, let A
k
j
: Y
k
(X
k
j
be a multimap with nonempty
convex values such that for each any x
k
j
X
k
j
, A
1
k
j
(x
k
j
) is open and

A
k
j
is an u.s.c.
multimap. For each k I and j J
k
, let the cost multimap F
k
j
: X
k
j
Y
k
( R
k
j
be
continuous with nonempty compact values and for any y
k
Y
k
, the multimap x
k
j
(
F
k
j
(x
k
j
; y
k
) is R
k
j
+
-quasiconvex. Then there exists y=( y
k
)
kI
Y, where y
k
=( x
k
j
)
jJ
k
,
with x
k
j

A
k
j
( y
k
) and there exist z
k
j
F
k
j
( x
k
j
; y
k
) such that
z
k
j
z
k
j
int R
k
j
+
;
for all z
k
j
F
k
j
(x
k
j
; y
k
), x
k
j

A
k
j
( y
k
k
.
Proof. For each k I and j J
k
and for all n N, we let a R
k
j
and e
k
j
int R
k
j
+
be
xed, and we dene H
k
j
;n
: Y
k
(X
k
j
by
H
k
j
;n
(y
k
) =
_
x
k
j
A
k
j
(y
k
): min p
k
j
F
k
j
(x
k
j
; y
k
) min p
k
j
F
k
j
(

A
k
j
(y
k
); y
k
) +
1
n
_
;
where the function p
k
j
: R
k
j
R is continuous and strict monotonically increasing
with respect to R
k
j
+
dened by
p
k
j
(z) = min{t R: z a + te
k
j
R
k
j
+
}
for all z R
k
j
.
Following the argument as in the proof of Theorem 4.1 and using Lemma 2.6, we
can show that for each k I and j J
k
and for all y
k
Y
k
and n N, H
k
j
;n
(y
k
) is a
nonempty convex set and for each x
k
j
X
k
j
, H
1
k
j
;n
(x
k
j
) is open in Y
k
.
Since for xed k I and for all n N and y
k
Y
k
, H
k
j
;n
(y
k
) is nonempty for all
j J
k
, we can let x
k
j
H
k
j
;n
(y
k
) for all j J
k
, then y
k
H
1
k
j
;n
(x
k
j
) for all j J
k
. There-
fore, y
k
jJ
k
H
1
k
j
;n
(x
k
j
) and we have
Y
k
=
_
_
_
_
jJ
k
H
1
k
j
;n
(x
k
j
): x
k
j
X
k
j
_
_
_
=
_
_
_
_
int
Y
k
jJ
k
H
1
k
j
;n
(x
k
j
): x
k
j
X
k
j
_
_
_
:
From Theorem 3.3, there exists
n
y = (
n
y
k
)
kI
Y, where
n
y
k
= (
n
x
k
j
)
jJ
k
, with
n
x
k
j

A
k
j
(
n
y
k
) such that
min p
k
j
F
k
j
(
n
x
k
j
;
n
y
k
) min p
k
j
F
k
j
(

A
k
j
(
n
y
k
);
n
y
k
) +
1
n
;
where
n
y
k
= (
n
y
l
)
lI; l=k
.
Let
n
a
k
j
F
k
j
(
n
x
k
j
;
n
y
k
) be such that p
k
j
(
n
a
k
j
)=min p
k
j
F
k
j
(
n
x
k
j
;
n
y
k
). Since for each
k I and j J
k
, X
k
j
is compact and F
k
j
and

A
k
j
are u.s.c. with compact values, it
follows from Lemma 2.2 that

A
k
j
(Y
k
) and F
k
j
(X
k
j
; Y
k
) are compact. Since for each
n N,
min p
kj
F
kj
(
n
x
k
j
;
n
y
k
) min p
k
j
F
k
j
(

A
k
j
(
n
y
k
);
n
y
k
) +
1
n
;
where
n
y
k
= (
n
y
l
)
lI; l=k
.
Let
n
a
k
j
F
k
j
(
n
x
k
j
;
n
y
k
) be such that p
k
j
(
n
a
k
j
) = min p
k
j
F
k
j
(
n
x
k
j
;
n
y
k
). Since for
each n N,
n
x
k
j

A
k
j
(
n
y
k
) and
n
a
k
F
k
j
(X
k
j
; Y
k
), there exist a subsequence {
n()
a
k
j
}
of {
n
a
k
j
}, a subnet {
n()
x
k
j
} of {
n
x
k
j
} and a
k
j
F
k
j
(X
k
j
; Y
k
), x
k
j

A
k
j
(Y
k
) such that
n()
x
k
j
x
k
j
and
n()
a
k
j
a
k
j
.
By Lemma 2.3 and following the argument as in the proof of Theorem 4.1, we note
that the function y
k
min p
k
j
F
k
j
(

A
k
j
( y
k
); y
k
) is continuous. Since
p
k
j
(
n()
a
k
j
) min p
k
j
F
k
j
(

A
k
j
(
n()

y
k
);
n()
y
k
) +
1
n()
;
we have
p
k
j
( a
k
j
) 6min p
k
j
F
k
j
(

A
k
j
( y
k
); y
k
);
by letting n() . As in the proof of Theorem 4.1, we can show that x
k
j

A
k
j
( y
k
)
and a
k
j
F
k
j
(x
k
j
; y
k
). This implies that for each k I and j J
k
,
p
k
j
( a
k
j
) 6p
k
j
(a
k
j
);
for all a
k
j
F
k
j
(x
k
j
; y
k
) and x
k
j

A
k
j
( y
k
).
Since p
k
j
is strict monotonically increasing with respect to R
k
j
+
,
a
k
j
a
k
j
int R
k
j
+
;
for all a
k
j
F
k
j
(x
k
j
; y
k
), x
k
j

A
k
j
( y
k
) and each k I and j J
k
.
From Theorem 4.2, we can easily derive the following result.
Corollary 4.1. For each k I and j J
k
, let A
k
j
: Y
k
( X
k
j
be a multimap with
nonempty convex values such that for any x
k
j
X
k
j
, A
1
k
j
(x
k
j
) is open in Y
k
and

A
k
j
is
an u.s.c. multimap. For each k I and j J
k
, let the cost function F
k
j
: X
k
j
Y
k
R
be continuous and for any y
k
Y
k
, the function x
k
j
F
k
j
(x
k
j
; y
k
) be R
+
-quasiconvex.
k
)
kI
Y, where y
k
= ( x
k
j
)
jJ
k
, with x
k
j

A
k
j
( y
k
) such that
F
k
j
(x
k
j
; y
k
) F
k
j
( x
k
j
; y
k
) 0
for all x
k
j

A
k
j
( y
k
k
.
Remark 4.2. For each k I , let J
k
={k} be a singleton set. If for each k I , A
k
: X
k
(X
k
is dened by A
k
(x
k
)=X
k
for all x
k
X
k
and F
k
: X
k
X
k
R is a single-valued
function, then Corollary 4.1 is reduced to the Nash equilibrium theorem [14].
k
, let F
k
j
, A
k
j
be the same as in Theorem 4.1
and also let W
k
j
R
k
j
+
\{0}. Assume that
(i) for each k I and j J
k
, the multimap W
k
j
F
k
j
: X
k
j
Y
k
( R is continuous
with nonempty compact values,
(ii) for each y
k
Y
k
, the multimap x
k
j
(W
k
j
F
k
j
(x
k
j
; y
k
) is R
+
-quasiconvex.
Then there exists y=( y
k
)
kI
, where y
k
=( x
k
j
)
jJ
k
, with x
k
j

A
k
j
( y
k
) and there exists
z
k
j
F
k
j
( x
k
j
; y
k
) such that
z
k
j
z
k
j
int R
k
j
+
;
for all z
k
j
F
k
j
(x
k
j
; y
k
), x
k
j

A
k
j
( y
k
k
.
Proof. From Theorem 4.2, there exists y = ( y
k
)
kI
, where y
k
= ( x
k
j
)
jJ
k
, with x
k
j

A
k
j
( y
k
) and there exists z
k
j
F
k
j
( x
k
j
; y
k
) such that
W
k
j
z
k
j
W
k
j
z
k
j
;
for all z
k
j
F
k
j
(x
k
j
; y
k
), x
k
j

A
k
j
( y
k
k
. Since W
k
j
R
k
j
+
\
{0}, W
k
j
z 0 for all z 0. Therefore,
z
k
j
z
k
j
int R
k
j
+
;
for all z
k
j
F
k
j
(x
k
j
; y
k
), x
k
j

A
k
j
( y
k
k
.
Now we adopt the xed point approach to prove the existence of a solution of
equilibrium problem with nite number of families of players and nite number of
families of constraints on the strategy sets.
Proposition 4.1. Let F
k
j
, A
k
j
, A
k
, W
k
j
and S
W
k
be the same as in Theorem 4.1. For
each k I and j J
k
, assume that F
k
j
is a continuous multimap with nonempty
compact values. Then y = ( y
k
)
kI

kI
Y
k
is a xed point of M if and only if
y
k
A
k
( y
k
k
j
F
k
j
( x
k
j
; y
k
) such that
W
k
j
z
k
j
6W
k
j
z
k
j
;
for all z
k
j
F
k
j
(x
k
j
; y
k
), x
k
j

A
k
j
( y
k
k
, where M is dened
in the beginning of this section.
Proof. Suppose that y is a xed point of M. Then for each k I , y
k
M
W
k
( y
k
).
Therefore, y
k
A
k
( y
k
) and min S
W
k
( y
k
; y
k
) =min S
W
k
(A
k
( y
k
); y
k
). This implies that
for each k I ,
min
jJ
k
W
k
j
F
k
j
( x
k
j
; y
k
) 6min
jJ
k
W
k
j
F
k
j
(x
k
j
; y
k
);
for all y
k
= (x
k
j
)
jJ
k
A
k
( y
k
). Therefore, for each k I and j J
k
jJ
k
min W
k
j
F
k
j
( x
k
j
; y
k
) 6
jJ
k
min W
k
j
F
k
j
(x
k
j
; y
k
); for all x
k
j
A
k
j
( y
k
):
(4.3)
Fix any k I j J
k
, let y
k
=(x
k
j
)
jJ
k
A
k
( y
k
) and x
ks
= x
ks
for s J
k
, s = j in (4.3).
Then (4.3) becomes,
min W
k
j
F
k
j
( x
k
j
; y
k
) 6min W
k
j
F
k
j
(x
k
j
; y
k
);
for all x
k
j
A
k
j
( y
k
k
. This implies that there exists
z
k
j
F
k
j
( x
k
j
; y
k
) such that
W
k
j
z
k
j
6W
k
j
z
k
j
;
for all z
k
j
F
k
j
(x
k
j
; y
k
), x
k
j
A
k
j
( y
k
k
.
Conversely, let y = ( y
k
)
kI
, where y
k
= ( x
k
j
)
jJ
k
A
k
( y
k
) and z
k
j
F
k
j
( x
k
j
; y
k
) be
such that
W
k
j
z
k
j
6W
k
j
z
k
j
;
for all z
k
j
F
k
j
(x
k
j
; y
k
), x
k
j
A
k
j
( y
k
k
. Then for each
k I ,
jJ
k
min W
k
j
F
k
j
( x
k
j
; y
k
) 6
jJ
k
min W
k
j
F
k
j
(x
k
j
; y
k
);
for all y
k
=(x
k
j
)
jJ
k
A
k
( y
k
). Therefore, min S
W
k
( y
k
; y
k
) =min S
W
k
(A
k
( y
k
); y
k
) and
y
k
M
W
k
( y
k
) for each k I . Hence
y = ( y
k
)
kI

kI
M
W
k
( y
k
) = M( y):
This shows that y is a xed point of M.
Remark 4.3. If I ={1; 2}, then Proposition 4.1 is reduced to Theorem 3.1 in [8].
As a consequence of Proposition 4.1, we have the following result.
k
, let X
k
j
be a nonempty compact convex subset
of a locally convex topological vector space E
k
j
, W
k
j
R
k
j
+
\{0} and let F
k
j
: X
k
j
Y
k
(
R
k
j
be a multimap with nonempty values. For each k I , assume that
(i) the multimap S
W
k
is continuous with compact values,
(ii) for each j J
k
, the multimap A
k
j
: Y
k
(X
k
j
is continuous with nonempty closed
values,
(iii) for all y
k
Y
k
, M
W
k
(y
k
) is an acyclic set.
k
)
kI
Y, where y
k
= ( x
k
j
)
jJ
k
with x
k
j
A
k
( y
k
) and there
exists z
k
j
F
k
j
( x
k
j
; y
k
) such that
z
k
j
z
k
j
int R
k
j
+
;
for all z
k
j
F
k
j
(x
k
j
; y
k
), x
k
j
A
k
j
( y
k
k
.
Proof. Since for each k I and for all y
k
Y
k
, A
k
(y
k
) =
jJ
k
A
k
j
(y
k
) is a closed
subset of a compact set Y
k
, A
k
(y
k
) is compact. Since S
W
k
(; y
k
) is a continuous multi-
map with nonempty compact values, it follows from Lemma 2.2 that S
W
k
(A
k
(y
k
); y
k
)
is a compact set and there exists a u
k
A
k
(y
k
) such that
min S
W
k
(A
k
(y
k
); y
k
) = min S
W
k
(u
k
; y
k
);
that is, M
W
k
(y
k
) is nonempty for all y
k
Y
k
and for each k I . By Lemmas 2.3
and 2.4, the functions (u
k
; y
k
) min S
W
k
(u
k
; y
k
) and y
k
min S
W
k
(A
k
(y
k
); y
k
) are
continuous. It is easy to see that for each k I and for all y
k
Y
k
, M
W
k
(y
k
) is a closed
subset of Y
k
and M
W
k
: Y
k
(Y
k
is closed. Since Y
k
is compact, it follows from Lemma
2.2 and condition (iii) that M
W
k
: Y
k
(Y
k
is an u.s.c. multimap with compact acyclic
values. By Kunneth formula [13] and Lemma 3 of Fan [5], M =
kI
M
W
k
: Y (Y is
an u.s.c multimap with compact acyclic values, that is, M V(Y; Y). Then it follows
from Lemma 2.5 that M has a xed point and the conclusion follows from Proposition
4.1 with W
k
j
R
k
j
+
\{0} for each k I and j J
k
.
Corollary 4.2. Let I be an index set and for each k I , let X
k
be a nonempty compact
convex subset of a locally convex topological vector space E
k
and let A
k
: X
k
( X
k
be a continuous multimap with nonempty closed values. For each k I , assume that
the single valued function F
k
: X
k
X
k
R
k
is continuous with nonempty compact
values and for any x
k
X
k
, the function x
k
F
k
(x
k
; x
k
) is R
k
+
-convex. Then there
exists x = ( x
k
)
kI
with x
k
A
k
( x
k
) such that
F
k
(x
k
; x
k
) F
k
( x
k
; x
k
) int R
k
+
;
for each k I .
Proof. It is easy to see that all conditions of Theorem 4.4 are satised and the con-
clusion of Corollary 4.2 follows immediately from Theorem 4.4.
Remark 4.4. If for each k I and j J
k
, A
k
j
: Y
k
(X
k
j
is a continuous multimap with
nonempty closed convex values, then the condition (iii) in Theorem 4.4 is replaced by
the following condition:
(iii)
For each k I and for each xed y

k
Y
k
, the multimap u
k
( S
W
k
(u
k
; y
k
) is
R
+
-quasiconvex.
Proof. It is sucient to show that for each k I and for all y
k
Y
k
, M
W
k
(y
k
) is a
convex subset of Y
k
. Indeed, let u
k
, v
k
M
W
k
(y
k
) and [0; 1], then u
k
; v
k
A
k
(y
k
)
and there exist a S
W
k
(u
k
; y
k
), b S
W
k
(v
k
; y
k
) such that a = b = min S
W
k
(A
k
(y
k
); y
k
).
By condition (iii)
, for each y
k
Y
k
and each R, the set
H() ={y
k
Y
k
: there is a c S
W
k
(y
k
; y
k
) such that c 6}
is convex. Take = min S
W
k
(A
k
(y
k
); y
k
), we see that u
k
; v
k
H(). Therefore, u
k
+
(1)v
k
H(). This implies that there is a c S
W
k
(u
k
+(1)v
k
; y
k
) such that c 6
= min S
W
k
(A
k
(y
k
); y
k
). Since for each k I , A
k
(y
k
) is convex, u
k
+ (1 )v
k

A
k
(y
k
) and
min S
W
k
(u
k
+ (1 )v
k
; y
k
) = min S
W
k
(A
k
(y
k
); y
k
):
This implies that for each k I , u
k
+ (1 )v
k
M
w
k
(y
k
) and M
W
k
(y
k
) is
convex. Hence for each k I , M
W
k
(y
k
) is acyclic and the conclusion follows from
Theorem 4.4.
Remark 4.5. In Theorem 4.1, if we assume that for each k I and j J
k
, X
k
j
is a
nonempty compact convex subset of a locally convex topological vector space E
k
j
,
then Theorem 4.1 is a simple consequence of Theorem 4.4 along with Remark 4.4.
Proof. Since for each k I and j J
k
, A
k
j
is l.s.c., it is easy to see that

A
k
j
: Y
k
(X
k
j
is l.s.c. and since

A
k
j
: Y
k
( X
k
j
is u.s.c., we have

A
k
j
: Y
k
( X
k
j
is continuous. Thus
the conclusion follows from Theorem 4.4 along with Remark 4.4.
Remark 4.6. If for each k I , F
k
: X
k
X
k
( R is a single-valued function, then
Corollary 4.2 is reduced to Debreu social equilibrium theorem [5].
5. Equilibrium problems with nite families of players and two nite families of
constraints on strategy sets
In this section, we establish existence results for a solution of the equilibrium prob-
lems with nite number of families of players and two nite number of families of
constraints on the strategy sets by using our coincidence Theorem 3.2 for two families
of multimaps.
k
, let X
k
j
be a nonempty compact convex subset
of a locally convex topological vector space E
k
j
and F
k
j
: X
k
j
Y
k
(R
k
j
be a multimap
with nonempty values. Let A
k
j
: Y
k
(X
k
j
be a multimap with nonempty convex values
such that for each x
k
j
X
k
j
, A
1
k
j
(x
k
j
) is open in Y
k
and

A
k
j
is an u.s.c. multimap. For
each k I , let B
k
: Y
k
(Y
k
be a closed multimap with nonempty convex values. For
each k I , assume that
(i) Y
k
=
{int
Y
k
B
1
k
(y
k
): y
k
Y
k
},
(ii) S
W
k
: Y
k
Y
k
(R is continuous multimap with compact values,
(iii) for any y
k
Y
k
, u
k
(S
W
k
(u
k
; y
k
) is R
+
-quasiconvex.
Then there exist y =( y
k
)
kI
Y, u =( u
k
)
kI
Y with y
k

A
k
( u
k
) and u
k

B
k
( y
k
)
and there exists z
k
j
F
k
j
( x
k
j
; u
k
) such that
z
k
j
z
k
j
int R
k
j
+
;
for all z
k
j
F
k
j
(x
k
j
; u
k
), x
k
j

A
k
j
( u
k
k
.
Proof. For each k I and for all n N, we dene H
k; n
: Y
k
(Y
k
by
H
k; n
(u
k
) =
_
y
k
A
k
(u
k
): min S
W
k
(y
k
; u
k
) min S
W
k
(

A
k
(u
k
); u
k
) +
1
n
_
:
As in the proof of Theorem 4.1, for all y
k
Y
k
, H
k; n
(y
k
) is a nonempty convex set
and
Y
k
=
_
{int
Y
k H
1
k; n
(u
k
): u
k
Y
k
}:
Since Y
k
=
{int
Y
k
B
1
k
(y
k
): y
k
Y
k
}, it follows from Theorem 3.2 that there exist
n
y = (
n
y
k
)
kI
Y =
jI
Y
j
and
n
u = (
n
u
k
)
kI
Y such that
n
y
k
H
k; n
(
n
u
k
) and
n
u
k
B
k
(
n
u
k
):
Therefore,
n
y
k
A
k
(
n
u
k
) and
min S
W
k
(
n
y
k
;
n
u
k
) min S
W
k
(

A
k
(
n
u
k
);
n
u
k
) +
1
n
:
Let
n
a
k
S
W
k
(
n
y
k
;
n
u
k
) be such that
n
a
k
=min S
W
k
(
n
y
k
;
n
u
k
). As in the proof of Theo-
rem 4.1,

A
k
(Y
k
) and S
W
k
(

A
k
(Y
k
); Y
k
) are compact and thus there exist a subnet {
n()
y
k
}
of {
n
y
k
}, a subsequence {
n()
a
k
} of {
n
a
k
} and y
k

A
k
(Y
k
), a S
W
k
(

A
k
(Y
k
); Y
k
) such
that
n()
( y
k
) y
k
and
n()
a
k
a
k
:
Since for each k I , B
k
: Y
k
(Y
k
is closed and Y
k
is compact, it follows from Lemma
2.2 that B
k
: Y
k
(Y
k
is an u.s.c. multimap with compact values and B
k
(Y
k
) is compact.
Therefore, there exist a subnet {
n()
u
k
} of {
n
u
k
} and u
k
B
k
(Y
k
) such that
n()
u
k
u
k
:
As in the proof of Theorem 4.1, for each k I ,

A
k
: Y
k
( Y
k
is a continuous multi-
map with compact values. By Lemma 2.3, the function u
k
min S
W
k
(

A
k
(u
k
); u
k
) is
continuous for each k I . Since
n()
a
k
min S
W
k
(

A
k
(
n()
u
k
);
n()
u
k
) + 1=n(), letting
n() , we have
a
k
6min S
W
k
(

A
k
(u
k
); u
k
); for each k I:
As in the proof of Theorem 4.1, we see that for each k I , a
k
S
W
k
(

A
k
( u
k
); u
k
),
y
k
= ( x
k
j
)
jJ
k

A
k
( u
k
) and u
k
B
k
( y
k
). Therefore, for each k I and j J
k
, there
exist x
k
j

A
k
j
( u
k
), z
k
j
F
k
j
( x
k
j
; u
k
) such that
a
k
=
jJ
k
W
k
j
z
k
j
6
jJ
k
W
k
j
z
k
j
; (5.1)
for all z
k
j
F
k
j
(x
k
j
; u
k
), x
k
j

A
k
j
( u
k
). For each xed k I and xed j J
k
, let y
k
=
(x
hs
)
sJ
k

A
k
( u
k
) and z
k
=(z
ks
)
sJ
k

sJ
k
F
ks
(x
k
ks
; u
k
) with x
ks
= x
ks
and z
ks
= z
ks
for
s J
k
, s = j, in (5.1), we have
W
k
j
z
k
j
6W
k
j
z
k
j
;
for all z
k
j
F
k
j
(x
k
j
; u
k
), x
k
j

A
k
j
( u
k
k
. Therefore,
z
k
j
z
k
j
int R
k
j
+
;
for all z
k
j
F
k
j
(x
k
j
; u
k
), x
k
j

A
k
j
( u
k
k
.
Theorem 5.2. Under the assumption of Theorem 4.2, we further assume that for
each k I , B
k
: Y
k
( Y
k
is a closed multimap with nonempty convex values and
Y
k
=
{int
Y
k
B
1
k
(y
k
): y
k
Y
k
}. Then there exist y = ( y
k
)
kI
Y, u = ( u
k
)
kI
Y
with y
k

A
k
( u
k
) and u
k
B
k
( y
k
k
j
F
k
j
( x
k
j
; u
k
) such that
z
k
j
z
k
j
int R
k
j
+
;
for all z
k
j
F
k
j
(x
k
j
; u
k
), x
k
j

A
k
j
( u
k
k
, where y
k
=( x
k
j
)
jJ
k
.
Proof. For each k I and j J
k
, let H
k
j
;n
: Y
k
( X
k
j
and p
k
j
: Z
k
j
R be dened as
in the proof of Theorem 4.2 and let
G
k; n
(y
k
) =
jJ
k
H
k
j
;n
(y
k
):
As in the proof of Theorem 4.2, we can show that Y
k
=
{int G
1
k; n
(y
k
): y
k
Y
k
}.
Since Y
k
=
{int B
1
k
(y
k
): y
k
Y
k
}, it follows from Theorem 3.2 that there exist
n
y =(
n
y
k
)
kI
Y and
n
u =(
n
u
k
)
kI
Y such that
n
y
k
G
k; n
(
n
u
k
) and
n
u
k
B
k
(
n
y
k
).
Therefore, for each k I ,
n
y
k
= (
n
x
k
j
)
jJ
k
A
k
(
n
u
k
)
and
min p
k
j
F
k
j
(
n
x
k
j
;
n
u
k
) min p
k
j
F
k
j
((

A
k
j
(
n
u
k
);
n
u
k
);
n
u
k
) +
1
n
:
where
n
y
k
= (
n
y
j
)
jI; j=k
. Following the arguments of Theorems 4.2 and 5.1, we get
the conclusion.
Acknowledgements
The authors are grateful to the referee for the careful reading and for her/his valuable
suggestions to make the paper in the present form.
References
[1] Q.H. Ansari, A. Idzik, J.C. Yao, Coincidence and xed point theorems with applications, Top. Meth.
Nonlinear Anal. 15 (2000) 191202.
[2] J.-P. Aubin, Mathematical Methods of Game and Economic Theory, North-Holland Publishing Company,
Amsterdam, 1979.
[3] J.-P. Aubin, A. Cellina, Dierential Inclusions, Springer, Berlin, 1994.
[4] H. Ben-El-Mechaiekh, P. Deguire, A. Granas, Une Alternative non lin eare en analyse convexe
et applications, C.R. Acad. Sci. Paris S er. I: Math. 295 (1982) 257259.
[5] G. Debreu, A social equilibrium theorem, Proc. Nat. Acad. Sci. U.S.A. 38 (1952) 386393.
[6] K. Fan, Fixed point and minimax theorems in locally convex topological linear spaces, Proc. Nat. Acad.
Sci. U.S.A. 38 (1952) 121126.
[7] B.S. Lee, G.M. Lee, S.S. Chang, Generalized vector variational inequalities for multifunctions,
Proceedings of Workshop on Fixed Point Theory, Vol. LI. 2, Universytet Marii Curie-Sklodowskilj,
Lublin, 1997, pp. 193202.
[8] L.J. Lin, On the systems of constrained competitive equilibrium theorems, Nonlinear Anal. 47 (2001)
637648.
[9] L.J. Lin, S. Park, On some generalized quasi-equilibrium problems, J. Math. Anal. Appl. 224 (1998)
167181.
[10] L.J. Lin, Z.Y. Yu, On some equilibrium problems for multimaps, J. Comput. Appl. Math. 129 (2001)
171183.
[11] D.T. Luc, Theory of Vector Optimization, in: Lecture Notes in Economics and Mathematical Systems,
Vol. 319, Springer, Berlin, 1989.
[12] D.T. Luc, C. Vargas, A saddle point theorem for set-valued maps, Nonlinear Anal. Theory Meth. Appl.
18 (1992) 17.
[13] W.S. Massey, Singular Homology Theory, Springer, Berlin, 1980.
[14] J. Nash, Noncooperative games, Ann. Math. 54 (1951) 286295.
[15] S. Park, Some coincidence theorems on acyclic multifunctions and applications to KKM theory, in:
K.K. Tan (Ed.), Fixed Point Theory and Applications, World Scientic, Singapore, 1991, pp. 248277.
[16] K.K. Tan, J. Yu, X.Z. Yuan, Existence theorems for saddle points of vector valued maps, J. Optim.
Theory Appl. 89 (1996) 731747.
[17] S.Y. Wang, An existence theorem of a Pareto equilibrium, Appl. Math. Lett. 4 (1991) 6163.
[18] S.Y. Wang, Existence of a Pareto equilibrium, J. Optim. Theory Appl. 79 (1993) 373384.
[19] J. Yu, G.X-Z. Yuan, The study of Pareto equilibria for multiobjective games by xed point and Ky
Fan minimax inequality methods, Comput. Math. Appl. 35 (1998) 1724.
[20] X-Z. Yuan, E. Tarafdar, Non-compact Pareto equilibria for multiobjective games, J. Math. Anal. Appl.
204 (1996) 156163.
[21] E. Zeidler, Nonlinear Functional Analysis and Its Applications I, Fixed Point Theorems, Springer, Berlin,
1985.

On The Constrained Equilibrium Problems With "Nite Families of Players

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

On The Constrained Equilibrium Problems With "Nite Families of Players

Hochgeladen von

Copyright:

Verfügbare Formate

Nonlinear Analysis 54 (2003) 525543

Corresponding author. Tel.: +886-4-7232105 ext. 3219; fax: +886-4-7211192.

U(x); (ii) lower semicontinuous (l.s.c.) if

V(x); (iii) continuous if

be a topological vector space. Let

be a family of nonempty convex subsets with each X

denotes the th coordinate and x

the projection of x onto X

and we also write

) is a nonempty convex set;

and a nonempty compact

such that for each x

be a nonempty compact convex subset of a

) is a nonempty convex subset of X

) is a nonempty convex subset of X

. Similarly, it follows from conditions (c), (d) and

has a continuous selection g

, by Tychono xed point theorem, there exists

)), that is, for each , x

)). For each , let u

For each k I and for each xed y

Das könnte Ihnen auch gefallen