Homoclinic Points and Homoclinic Orbits for the Quadratic Family of Real Functions with Two Parameters ()

1. Introduction
Recently, we witness that there is a clear attention in connecting orbits (homoclinic or hetroclinic orbits) in dynamical systems [1] . The notion of a homoclinic point was first introduced by ponicaré (1890) [2] , in the study of a three-body problem. After about 70 years, Smale (1963, 1976) [3] , presented the horseshoe notion to show the chaotic behavior of higher-dimensional systems with a transverse homoclinic point. Homoclinic points have been used to study the dynamics of two- or higher-dimensional dynamical systems. Devaney (1989) in [4] , showed that the same result holds for a one-dimensional map of an interval into itself with a nondegenerate homoclinic point. Block and Coppel (1992) [5] , proved that for one-dimensional map f of an interval into itself, a homoclinic point leads to horseshoe for
. Recently, Li (2002) gave a simple proof of Block and Coppel result proof. The concept of homoclinic orbits and heteroclinic connections plays a central role in the studying the chaotic sets. In particular, it can be used for proof of the existence of chaos. In fact, it is proved in [6] [7] [8] , the existence of a non-degenerate homoclinic orbit to an expanding fixed point of a smooth map f implies the existence of an invariant set in a neighborhood of the homoclinic orbit, on which f is chaotic. The same result is also true for non-degenerate heteroclinic connections. Homoclinic orbits and heteroclinic connections are relevant not only for the proof of the existence of chaos but also for the description of several bifurcations of chaotic attractors (crisis bifurcations) [9] . In this work, we find homoclinic points and homoclinic orbits for two-parameters family of real functions
. In Section 2, we study the fixed points of the family
where
, and the nature of this fixed point for various values a and b. In Section 3, we study the local unstable sets of the repelling fixed point
for the functions
. We proved
that
. And in Section 4, we study the unstable sets of the repelling fixed points
for the functions
we proved that
. Finally, in Section 5, we study the homoclinic points and homoclinic orbits of the function
for the repelling fixed point
. We show that
has a homoclinic point and orbit whenever
, and has no homoclinic point and orbit whenever
. We need some important definitions.
Definition 1: Let P be a repelling fixed point for a map
on a compact interval
. Then the unstable set of P is defined as
[10] , which equivalent to the expansivity
[4] .
Definition 2: Let P be a repelling fixed point for a map
on a compact interval
and U be an open interval near P. Then, the local unstable set
of P is defined as
[10] . Which equivalent to the expansivity
[11] ,
and equivalent to
for any
[6] .
Definition 3: Let P is fixed point and
for a map
on a compact interval
. A point q is called homoclinic pint to P if
and there exists
such that
[4] .
Note that the sequence of images of a homoclinic point q and a suitable sequence of preimages of q consist of points which are also homoclinic, and both these sequences converge to P. The union of these sequences
where
for
,
and
, is called the homoclinic orbit of P [12] .
2. The Fixed Points of the Function
and Their Nature
In this section we study the fixed points of the family
where
, and the nature of these fixed points for various values a and b. It is clear that the fixed points of
are
and
. The graphs show that the function
has no fixed point for
, has a unique indifferent fixed point for
and has two fixed points for
. See Figure 1.
The fixed point
is always repelling for
. But the fixed point
is attracting for
, indifferent for
and is repelling for
. See Figure 2.
We need the following remarks in our work.
2.1. Remark
For
,
and for
,
.
Proof:
It is clear that
, so for
,
. If
, then
, is defined and
. Now for the fixed point
,
. If
, then
, thus
, there for
According to the definition of the homoclinic points, we consider the repelling fixed point
and we omit the repelling fixed point
.
2.2. Remark
For
with
, the fixed point
.
Proof:
Let
. Then
is defined and thus,
. There for
![]()
Figure 1. (a) a = 1, b = 0.5; (b) a = 1, b = −0.25; (c) a = 1, b = −1.
3. The Local Unstable Sets
In this section, we study the local unstable sets of the repelling fixed point
for the functions
.
Proposition
For
the local unstable set of the fixed point
is
.
Proof:
To find the local unstable for the repelling fixed point
we consider the inequality
[6] . Since
is a continuous function, then there exists a neighborhood U of the fixed point
such that
, i.e.
. Thus
. Then either
, i.e.
, or
, i.e.
. By remark (2.2)
, there for
.
4. The Unstable Sets
The unstable sets of the repelling fixed points
for the functions
is calculated in the following proposition.
We need the following lemma in our studying.
4.1. Lemma
For
,
where
.
The proof is clear.
4.2. Proposition
For
the unstableset of the fixed point
is
.
Proof:
Consider the expansive inequality
where
.
Then for
, we have
, thus
. By lemma (4.1). Thus we have
. There for
.
Either
which implies
. Or
, which implies
. That is
. To calculate
, by remark (2.2),
. Now
, which implies
. So
.
4.3. Example
For
, then
.
Solution:
It is clear that
. Since
, where
, implies that
, then
. Thus
. Since
, then
, there for either
, then
, i.e.
.
Or
, then
, i.e.
. Thus
.
But
, then
.
5. Homoclinic Points and Homoclinic Orbits for the Family H
In this section, we study the homoclinic points and homoclinic orbits of the function
for the repelling fixed point
where
, note that the other repelling fixed point
. By remark (2.1),
where
. Thus according to the definition of the homoclinic point, we don’t study the homoclinic points of
.
5.1. Homoclinic Points for the Functions
To study the homoclinic points of the repelling fixed point
of
we use the following technique concerned on the preimages of
.
5.1.1. Remark
For
,
.
Proof:
By lemma (4.1),
. Then
. For
, so we have the same state, so is omitted. For
.
Now if
, in remark (5.1.1), then put
where
. For
, where
. Similarly for
, where
.
And so on (see Figure 3). Thus we have sequences of preimage points of
,
. If
, for some
, then these points are the homoclinic points of the repelling fixed point
.
In following propositions and examples we study the homoclinic points of the fixed point
of
for various values of a and b.
![]()
Figure 3. Tree of the homoclinic points of
.
5.1.2. Proposition
For
. If
then the fixed point
has no homoclinic points.
Proof:
It is clear that
and
. By proposition (3.1),
. Now the first preimage of
is
where
. For
and by lemma (4.1), we have
. But
is a fixed point, and
. Moreover, by proposition (5.1.1), we have
, since
. There for
is undefined in
for
. Thus
has no homoclinic point to the fixed point
.
5.1.3. Example
For
has no homoclinic points to the fixed point
.
Solution:
By proposition (3.1),
. The first preimage of
is
where
. For
. But
is a fixed point of
, and
. Now for
. So
has no homoclinic points to the fixed point
.
5.1.4. Proposition
For
, the first preimage of the fixed point
cannot be a homoclinic point to the fixed point
.
Proof:
The first preimage of
is
where
. For
and by lemma (4.1) we have
where
. But
is a fixed point, by remark (2.2)
, we have
and
.
So
. Thus the fixed point
has no homoclinic points at the first preimage.
5.1.5. Example
has no homoclinic points to the fixed point
in the first preimage.
Solution:
The fixed point
, and the first preimage of
is
. From proposition (3.1),
. For
then
. But
is a fixed point of
and
. So
has no homoclinic points to the fixed point
in the first preimage.
Following remarkes assert that some inverse images of
,
are belong to the local unstable set of
for
.
5.1.6. Remark
If
of
, then the second preimage of the fixed point
belongs to the local unstable set of
.
Proof:
Let
, it is clearly that
. There for
and
. So
. Then
, thus
, which implies
. Since
, there for
, then
. Thus
, i.e.
. There for
. By remark (5.1.1)
, So
. Thus
. (See proposition (3.1)).
5.1.7. Remark
If
of
, then the third preimage of the fixed point
belongs to the local unstable set of
.
The proof is same of the above remark with some more complicated details.
5.1.8. Theorem
For the family
, there exist homoclinic points to the fixed point
whenever
.
Proof:
According to a proposition (5.1.4), we begin with the second preimage of the fixed point
. By remark (5.1.1), we have
. Suppose that
, i.e.
, then
. Since
, where
. Then
, which implies
.
Let
. Since
, then
, which implies
, i.e.
………(*).
Then we have three cases for b
Case 1: If
, then
, thus
, there for
, which implies
which is a contradiction with
in (*).
Case 2: If
, then by proposition (5.1.2) there is no homoclinic points to the fixed point
.
Case 3: If
, then
. Thus
, there for
, which implies
, i.e.
. Since
in (*), then
. There for
, implies that
, thus
………(**).
Since
and
, then
, so by (**),
, thus
. It follows that
. By remark (2.1),
.
Now:
1) For
, let
, by remark (5.1.6),
and it is clear that
(see Figure 3).
2) For
, let
, by remark (5.1.7),
and it is clear that
(see Figure 3).
There for
for
and
for
are the first homoclinic points for the repelling fixed point
.
5.1.9. Theorem
If
, then
has no homoclinic points to the fixed point
.
Proof:
According to a proposition (5.1.4), we begin with the second preimage of the fixed point
. By remark (5.1.1), we have
. Since
is a repelling fixed point for
. So, with our assumption, we have
. Now we divide the proof into three cases:
Case 1: If
, then
. Thus
, and
. So
, which implies
. Thus
……… (*).
Since
and
, then
. Thus (*) will being
, which implies,
, i.e.
, which implies
. So
. There for
. Then
has no homoclinic points to the fixed point
for
.
Case 2: If
, then by proposition (5.1.2) there is no homoclinic point to the fixed point
.
Case 3: If
, then
. Thus
, and
. So
, which implies
, thus
……… (*), Since
and
, then
. Thus (*) will being
, which implies,
, i.e.
, which implies
. So
. But by theorem (5.1.8), if
, then
which is a contradiction with
. There for
has no homoclinic points to the fixed point
for
.
Following examples explain the cases for
,
and
respectively.
5.1.10. Example
For
, has a homoclinic point to the fixed point
at (0.86602540378).
Solution:
The fixed point
and the first preimage of
is
. From proposition (3.1),
. Clearly
. For
, then
. But
is a fixed point of
, and
. Now for
,
where
and
. Moreover
and
So (0.86602540378) is a homoclinic point to the fixed point
.
5.1.11. Remark
Here we consider 0.86602540378 (the first) homoclinic point for the fixed point (1.5). In fact, there are many points belong to the local unstable set of 1.5 (i.e. homoclinic points to
). In fact
.
Now,
, and
. For
,
and
, and
. If we continue with this way, we get a set
. Every point in this set belongs to
. Each point of this set is a homoclinic point to the fixed point
. See Figure 4.
5.1.12. Example
For
,
is a homoclinic point to the fixed point
.
Solution:
It is clear that
, and the first preimage of
is
.
From proposition (3.1),
. Clearly
. For
, then
. But
is a fixed point of
, and
. Now for
. Now for
.
and
. Moreover
,
and
. So
is a homoclinc point to the fixed point
.
In fact, any preimage point contained in
is a homoclinic point for
. See Figure 5.
5.1.13. Example
, has no homoclinic points to the fixed point
.
![]()
Figure 4. Tree of the homoclinic points of
.
![]()
Figure 5. Tree of the homoclinic points of
.
Solution:
It is clear that
, and the first preimage of
is
.
From proposition (3.1),
. For
, then
. But
is a fixed point of
, and
. Now for
. So
has no homoclinic points to the fixed point
5.2. Homoclinic Orbits for
In this part we study the homoclinic orbits for the family
.
5.2.1. Remark
For
, we proved in theorem (5.1.9),
has no homoclinic points, so
has no homoclinic orbits for the repelling fixed point
.
To study the homoclinic orbits of
for
, we introduce the following theorems and lemmas.
5.2.2. Lemma
For
with
, if
where
is a constant and q is a homoclinic point to
, then
.
Proof:
Let
, then
, which implies
, thus
.
Now since
, then
(see proposition (3.1)). So it is clearly
, thus
.
5.2.3. Lemma
, where
is a constant.
Proof:
Since
and
.
5.2.4. Theorem
Let
for
. If
, then
is an increasing sequence.
Proof:
The first preimage of
is
.
Claim:
is increasing for
. To show this, let
,
that is
, thus
, there for
. Then
either
and
, or
and
(omitted because there is no intersection). Hence
is increasing whenever
. Now
. Which implies
, (
is increasing). Thus, we have
, for any
.
5.2.5. Theorem
Let
and
. If
, then
is a decreasing sequence.
The proof is the same as the above theorem.
5.2.6. Theorem
For
with
, if
then the upper bound of the increasing sequence of preimages of x,
is
.
Proof:
It is clear that
. We will prove
by induction
Since
, then
. Since
and
(because
so
is undefined if
), then
. There for
. So, by lemma (4.1) then
……….(*).
Now since
, i.e.
. Then by (*) we have
, thus
. Thus by lemma (4.1), then
.
Now assume that
is true and we have to show
.
Since
. Then, with our assumption
, we get
, thus
. So by lemma (4.1),
.
5.2.7. Theorem
For
with
, if
then the lower bound of the decreasing sequence
is
.
The proof is the same as the above theorem.
5.2.8. Theorem
If
,
and,
then the supremum of the increasing sequence
is
.
.
Proof:
The first preimage of
is
.
If
. Let
and
. There for
. By lemma (5.2.2)
, there for
, which implies
.
Now let
. Then
such that
, thus
, which implies
, thus
. By lemma (4.1) and lemma (5.2.3),
……(*).
But
and
, which implies
, thus
, hence
, there for
. Thus by (*),
which is a contradiction with
. Thus
.
5.2.9. Theorem
For
of the functions
and
then the infimum of a decreasing sequence
is
.
.
The proof is the same as the above theorem.
Finally, we introduce the main theorem in this section.
5.2.10. Theorem
For
with
and
, the homoclinic orbist of the (first) homoclinic points
and
are
for
and
for
.
Proof:
Since
. Then
1) By Remark (2.1), for
,
.
2) By theorem (5.1.8),
for
and
for
(see Figure 3). So
for
,
for
.
3) By theorems (5.2.4) and (5.2.8), (resp. (5.2.5) and (5.2.9)),
where
is an increasing with supremum (resp. decreasing with infimum)
. Thus,
. There for, 1, 2 and 3 show that
,
are the homoclinic points for P1 with the homoclinic orbits
for
and
for
. See Figure 3.
Following examples explain the cases for
and
respectively.
5.2.11. Example
For
, a homoclinic orbit of a homoclinic point
is
.
Solution:
The forward orbit of
,
, thus
. So
for
.
The backward orbit of
is
.
To prove
. Consider the sequence
.
is an increasing sequence: It is easily shown that for
the function
is an increasing function. Hence
, thus
is an increasing sequence.
Moreover for
, by theorem (5.2.6)
(i.e. 2 is an upper bound for
). Thus to show that
converges to 2, it is enough prove that
. If
, let
and
. By lemma (5.2.2) then
, since
is an increasing sequence and
, then
. Now, since
, then
. Let
.
Then
such that
.
Since the iteration of this sequence is
, thus
, which implies
, then
………(*).
But
and
which implies
, therefor
. Thus by (*),
which is a contradiction with
. Thus
, and
. So
is a homoclinic orbit of a homoclinic point
for
.
5.2.12. Example
For
, a homoclinic orbit of a homoclinic point
is
.
Solution:
The forward orbit of
,
. So
for
.
The backward orbit of
is
.
To prove
. Consider the sequence
.
is an increasing sequence: It is easily shown that for
the function
is an increasing function. Hence
, thus
is an increasing sequence.
Moreover, for
, by theorem (5.2.6)
(i.e. 3 is an upper bound for
). Thus to show that
converges to 3, it is enough prove that
. If
, let
and
. By lemma (5.2.2) then
, since
is an increasing sequence and
, then
. Now, since
, then
. Let
.
Then
such that
.
Since the iteration of this sequence is
, thus
, which implies
, then
………(*).
But
and
which implies
, there for
. Thus by (*),
which is a contradiction with
. Thus
, and
. So
is a homoclinic orbit of a homoclinic point
for
.