Existence Theory and Properties of Solutions: TH (N) (N TH

Chapter 2
Existence Theory and Properties of

Solutions
This chapter contains some of the most important results of the course. Our rst goal is to
prove a theorem that guarantees the existence and uniqueness of a solution to an initial value
problem on some, possibly small, interval. We then investigate the issue of how large this
interval might be. The last section of the chapter provides some insight into how a solution
of an initial value problem changes when the dierential equation or initial conditions are
altered.
2.1 Introduction
Consider an n
th
order dierential equation in the form
y
(n)
= g(t, y, y
, y
, , y
(n1)
).
It is a standard practice to convert such an n
th
order equation into a rst order system by
dening
x
1
= y
x
2
= y
.
.
.
x
n
= y
(n1)
.
We will denote vectors in R
n
by x = (x
1
, x
n
) so that our scalar equation is now represented
in vector form as
dx
dt
= x
(t) =
_
_
_
_
_
x
2
x
3
.
.
.
g(t, x
1
, x
2
, , x
n
)
_
_
_
_
_
= f(t, x(t)).
1
2 CHAPTER 2. EXISTENCE THEORY AND PROPERTIES OF SOLUTIONS
Consequently it suces to focus upon 1
st
ordinary dierential equations denoted by
x
(t) = f(t, x(t)) (2.1.1)

where x R
n
and f(t, x) R
n
is dened on an open set U RR
n
. A solution of (2.1.1)
is a dierentiable function
: J R
n
where J is an open interval in R and for t J, (t, (t)) U, and
(t) = f(t, (t)).

A solution (t) of the initial value problem (IVP)
x
(t) = f(t, x(t))

x(t
0
) = x
0
.
(2.1.2)
is a solution of the dierential equation (2.1.1) that also satises the initial condition (t
0
) =
x
0
.
Example 2.1.1 Rcall from Example (1.2.3) that the IVP
x
=
x
t
+ t = f(t, x)
x(0) = x
0
has innitely many solutions if x
0
= 0 and no solution if x
0
= 0. This suggests that continuity
of f(t, x) would be a minimal condition to ensure existence of a solution to an IVP.
Example 2.1.2 Consider
x
= f(t, x) = x
1/3
.
By separation of variables we get the family of solutions
(t) = (
2
3
(t + c))
3/2
Now consider the IVP
x
= x
1/3
x(0) = 0.
2.1. INTRODUCTION 3
For each c > 0 we obtain a solution
c
where
c
(t) =
_
(
2
3
(t c))
3/2
, t c
0, t c.
Thus we see that continuity of f(t, x) is not enough to ensure uniqueness.
Fig. 2.1.1. There are innitely many solutions to the IVP in Example 2.1.2.
Our goal is to prove that under appropriate hypotheses on f(t, x), the initial value prob-
lem (2.1.2) has a solution dened on an interval (t
0
, t
0
+) and any two such solutions must
agree on their common domain. The above examples suggest that an appropriate notion of
smoothness must be assumed of f(t, x). To describe the regularity that will be required we
need to introduce some terminology. For x R
n
we denote the sup, or l
norm by
|x| = max
n
{|x
n
|}.
Let (X, d) be a metric space and denote the open ball of radius r around x
0
by
B
r
(x
0
) = {x | d(x, x
0
) < r}.
B
r
(x
0
) will denote the closed ball {x| d(x, x
0
) r}. Let J
(t) = (t , t +) R and assume

that f(t, x) : U RR
n
R
n
. The existence and uniqueness results we prove are obtained
by assuming that f(t, x) satises a Lipschitz condition. More precisely, we say that f(t, x)
is locally Lipschitz with respect to x if for any (t
0
, x
0
) U there exists L 0 and > 0, so
that J
(t
0
) B
(x
0
) U and
|f(t, x) f(t, y)| L|x y|, for t J
(t
0
) and x, y B
(x
0
).
It is easily veried that if f(t, x) is continuous and the partial derivatives f
i
/x
i
exist
and are continuous on U, then f is locally Lipschitz with respect to the second variable.
Previously we saw that the IVP of Example(2.1.2) has innitely many solutions. Note that
the function f(x) = x
1/3
is not Lipschitz at the origin.
The notion of a contractive mapping is central to many existence arguments. If statises
0 < < 1, and T : X X is a mapping, we say T is an -contraction if
d(T(x), T(y)) d(x, y) for all x, y X.
If T(p) = p, we call p is a xed point of T. We will denote iterates of T by
T
0
(x) = x, T
1
(x) = T(x), T
2
(x) = T(T
1
(x)), T
n
(x) = T(T
n1
(x)).
The following lemma is crucial.
Lemma 2.1.1 [Contraction Mapping Lemma]. Let (X, d) be a complete metric space
and T : X X an -contraction. Then T has a unique xed point p. In fact, for any
x X, the iterates T
n
(x) converge to p.
Proof. Dene f : X [0, ) by f(x) = d(T(x), x). In other words, f(x) is the distance
T moves x. Note that f(p) = 0 if T(p) = p and observe that f is continuous. Indeed
f(x) = d(x, T(x))
d(x, y) + d(y, T(y)) + (T(y), T(x))
d(x, y) + f(y) + d(x, y)
and so
f(x) f(y) (1 + )d(x, y).
Interchanging x and y we see
|f(x) f(y)| (1 + )d(x, y).
There are two inequalities satised by f. First,
f(T(x)) = d(T(T(x)), T(x)) d(T(x), x) = f(x). (2.1.3)
For the second inequality note that for x, y X,
d(x, y) d(x, T(x)) + d(T(x), T(y)) + d(T(y), y)
f(x) + d(x, y) + f(y)
and so
d(x, y)
f(x) + f(y)
1
. (2.1.4)
2.2. EXISTENCE AND UNIQUENESS OF SOLUTIONS 5
Now let x
0
be any point in X and x
n
= T
n
(x
0
). Then from (2.1.3)
f(x
n
)
n
f(x
0
)
and so f(x
n
) 0 as n . It follows from (2.1.4) that for any n, m
d(x
n
, x
m
)
f(x
n
) + f(x
m
)
1
.
For n, m suciently large we can make the right hand side as small as we like and hence
{x
n
} is a Cauchy sequence. Since X is complete, there exists a p X such that x
n
p.
Because f is continuous, f(x
n
) f(p), and so f(p) = 0, i.e., p is a xed point of T.
To show uniqueness suppose q is another xed point. Then f(q) = 0 and from (2.1.4) we
see d(p, q) = 0.
2.2 Existence and Uniqueness of Solutions
It turns out that continuity of f(t, x) is sucient to guarantee existence of a solution to the
IVP
x
(t) = f(t, x(t))

x(t
0
) = x
0
.
This result is referred to as Peanos Theorem. Example (2.1.2) in the previous section showed
that we need additional hypotheses on f(t, x) to ensure uniqueness. The condition we need
is Lipschitz continuity.
The next theorem is a rst form of our Existence and Uniqueness Theorem.
Theorem 2.2.1 Let f : U R R
n
R
n
, U open and f(t, x) continuous and locally
Lipschitz with respect to the second variable. The following two statements hold.
(1) Select (t
0
, x
0
) U. For all > 0 suciently small there is a dierentiable function
: (t
0
, t
0
+ ) R
n
such that
(t, (t)) U, t J
(t
0
)
(t) = f(t, (t)), t J
(t
0
)
(t
0
) = x
0
.
_
_
(2.2.1)
That is, is a solution of the initial value problem.
(2) If
1
: J
1
(t
0
) R
n
and
2
: J
2
(t
0
) R
n
are two dierentiable functions that satisfy
(2.2.1), then
1
and
2
agree on some open interval around t
0
.
First we need the following lemma.
Lemma 2.2.1 A function : J
(t
0
) R
n
is dierentiable and satises (2.2.1) if and
only if for t J
(t
0
), (t, (t)) U, is continuous, and satises
(t) = x
0
+
_
t
t
0
f(s, (s))ds, t J
(t
0
). (2.2.2)
Proof. Let be a dierentiable function that satises (2.2.1). Since
(t) = f(t, (t))

and f is continuous,
is continuous. Thus by the Fundamental Theorem of Calculus

_
t
t
0
(s)ds = (t) (t
0
) = (t) x
0
=
_
t
t
0
f(s, (s)) ds
and so (t) satises (2.2.2).
Conversely suppose satises the conditions of the second part of the lemma. Then
clearly
(t
0
) = x
0
and by the Fundamental Theorem of Calculus,
(t) = f(t, (t)).

Thus is dierentiable and satises the IVP.
The proof of Theorem(2.2.1) is based on the Contractive Mapping Lemma where the
underlying metric space will be a closed subset of BC(J
(t
0
); R
n
), the space of bounded
continuous functions
: J
(t
0
) R
n
where for
1
,
2
BC(J
(t
0
)),
d(
1
,
2
) = sup
tJ(t
0
)
{|(
1
2
)(t)|}
= ||
1
2
||.
Proof of Theorem 2.2.1. Since f is locally Lipschitz with respect to the second
variable, we can nd an r > 0 such that [t
0
r, t
0
+ r] B
r
(x
0
) U and
|f(t, x) f(t, y| L|x y| for all (t, x), (t, y) [t
0
r, t
0
+ r] B
r
(x
0
).
Since [t
0
r, t
0
+ r] B
r
(x
0
) is compact and f is continuous, there exists an M for which
|f(t, x)| M for all (t, x) [t
0
r, t
0
+ r] B
r
(x
0
).
Choose > 0 so small such that
< r
M < r
L < 1.
Let X BC(J
(t
0
); R
n
) be the space of continuous functions
: J
(t
0
) B
r
(x
0
).
Then X is a closed subset of BC(J
(t
0
); R
n
) and hence is complete. Note that if X, t
J
(t
0
) and < r, then we certainly have (t, (t)) J
(t
0
) B
r
(x
0
) U. For X, dene
T on J
(t
0
) by
T(t) = x
0
+
_
t
t
0
f(s, (s))ds.
By the Fundamental Theorem of Calculus T is continuous and
|T(t) x
0
| = |
_
t
t
0
f(s, (s))ds| |
_
t
t
0
|f(s, (s)|ds|
M|t t
0
| < M < r.
Hence T(t) B
r
(x
0
) and so T X. Thus T : X X.
We now show T is a contraction. If , X,
| T(t) T(t)| = | x
0
+
_
t
t
0
f(s, (s)) ds x
0
_
t
t
0
f(s, (s)) ds |
_
t
t
0
| f(s, (s)) f(s, (s)) | ds |
L|
_
t
t
0
| (s) (s) | ds |
L|
_
t
t
0
|| || ds |
= L|t t
0
| || ||
L|| ||.
Thus
sup
tJ(t
0
)
|T(t) T(t)| = ||T T|| L|| ||
and since L < 1, T is a contraction. Hence T has a xed point and so there exists X
such that
T(t) = (t) = x
0
+
_
t
t
0
f(s, (s))ds.
By Lemma(2.2.1), this (t) is a solution (2.2.1)
Before proceeding to the proof of the second statement, note that given (t
0
, x
0
) we rst
choose r such that K
r
= [t
0
r, t
0
+r] B
r
(x
0
) U. Once we select so that < r, M <
r, L < 1, we can consider the set X BC(J
(t
0
); R
n
) in which T has a xed point. In
this sense the set X may be regarded as a one-paremeter family X() and the xed point,
though unique in X(), does depend on .
To prove the second statement of the proposition suppose
1
,
2
are two solutions of the
IVP. The intersection of their domains is an open interval, say (t
0
, t
0
+). Since
1
(t
0
) =
2
(t
0
) = x
0
, and
1
,
2
are continuous we can select such that
1
,
2
: J
(t
0
) B
r
(x
0
). We
can further decrease if necessary so that < r, M < r and L < 1. With this choice
of , we then get that
1
,
2
X() and since T : X() X() has a unique xed point,
1
(t) =
2
(t), t J
(t
0
).
In summary, we have that for (t
0
, x
0
) U there exists > 0 such that the IVP has a
solution that is dened on (t
0
, t
0
+ ) if < r, M < r, L < 1. Note that in showing
T(t) B
r
(x
0
) we showed all iterates satisfy
|T x
0
| M|t t
0
| < M.
In particular the graph of the solution to (2.2.1) lies in the region R as depicted in the gure
below. Note that if M is large, the graph of the solution may escape the set K
r
unless the
domain of the solution is restricted as required by the condition M < r.
Fig. 2.2.1. The graph of T(t).
An improved statement of Theorem 2.2.1 constitutes our main Existence and Uniqueness
Theorem. Note that this a local result in that the time interval on which the solution exists
may be small.
Theorem 2.2.2 [Existence and Uniqueness] Assume f : U RR
n
R
n
is continuous
and locally Lipschitz with respect to the second variable. If (t
0
, x
0
) U, then there is an
> 0 such that the IVP
x
= f(t, x)
x(t
0
) = x
0
has a unique solution on the interval (t
0
, t
0
+ ).
Proof We know that for all suciently small , the initial value problem has a solution.
We need only prove that if
1
,
2
satisfy the IVP on J
(t
0
), then
1
=
2
on J
(t
0
).
Let S = {t (t
0
, t
0
+ )|
1
(t) =
2
(t)}. S is not empty since (t
0
) =
2
(t
0
). Since
1
and
2
are continuous, S is closed in J
(t
0
). Let
t S and x =
1
(
t) =
2
(
t)). Then
1
,
2
solve the IVP with initial condition (
t, x). By the previous proposition,

1
and
2
agree on
an open interval J S containing say

t. Hence S is open and closed and since J
(t
0
) is
connected, S = J
(t
0
).
The proof of Theorem(2.2.1) can be used to obtain a sequence of approximations that
converge to the solution of the IVP. It is customary to begin the iteration process with
x(t) = x
0
. Then
x
1
(t) = Tx = x
0
+
_
t
t
0
f(s, x
0
)ds
x
2
(t) = T
2
x = x
0
+
_
t
t
0
f(s, x
1
(s))ds
.
.
.
.
.
.
x
n
(t) = T
n
x = x
0
+
_
t
t
0
f(s, x
n1
(s))ds.
From our results we know that {x
n
(t)} converges to a solution of the IVP in some neigh-
borhood of t
0
. This sequence of approximate solutions are known as Picard iterates. The
usefulness of approximating a solution by this procedure has been somewhat enhanced by
the availability of computer algebra systems such and Maple and Mathematica.
2.3 Continuation and Maximal Intervals of Existence
Our existence theorem is of local nature in that it provides for the existence of a solution to
the IVP
x
(t) = f(t, x(t))

x(t
0
) = x
0
dened in an interval (t
0
, t
0
+ )
Example 2.3.1 The solution of
x
= x
2
x(0) = 1
is
x(t) =
1
1 t
Here U = RR. Note that the solution is dened for < t < 1. As t 1
, the graph of
the x(t) leaves every closed and bounded subset of U. We will prove a theorem that reects
this general behavior. That is, the solution of an IVP can be dened on an interval (m
1
, m
2
)
where either m
2
= +or the graph of the solution escapes every closed and bounded subset
of U as x m
2
(and similarly for m
1
).
2.3. CONTINUATION AND MAXIMAL INTERVALS OF EXISTENCE 11
Throughout this section we suppose f : U R R
n
R
n
, U open, and f(t, x) is
continuous on U and locally Lipschitz in x. Suppose (t) is a solution of
x
= f(t, x)
x(t
0
) = x
0
that is dened for < t < and (, (
)) U. Now consider the IVP

x
= f(t, x)
x() = (
).
We know this problem has a solution, say (t), dened on t < + . Dene
y(t) =
_
(t) < t <
(t) t < + .
Clearly y(t) is continuous. Moreover,
y(t) = (
) +
_
t
f(s, (s)) ds for t < +

and
(
) = x
0
+
_

x
0
f(s, (s))ds.
Hence
y(t) = x
0
+
_

t
0
f(s, (s)ds +
_
t
f(s, (s))ds
or
y(t) = x
0
+
_
t
t
0
f(s, y(s))ds, t < + .
Since we clearly have
y(t) = x
0
+
_
t
t
0
f(s, y(s))ds, < t < ,
it follows from Lemma(2.2.1) that
y
(t) = f(t, y(t)), < t < +

y(t
0
) = x
0
and so y(t) is a solution of the IVP that is dened on a larger interval.
The above process is referred to as continuation to the right. In the same way one could
construct a continuation to the left. By our uniqueness result any extension of the solution
from (, ) to (
1
, +
2
) is unique. The geometric interpretation of the continuation
process is displayed in Figure 2.3.1.
Fig. 2.3.1. The continuation process.
Denition (2.3.1) Let be a solution of an ordinary dierential equation on an interval J.
A function

is called a continuation of if
i)

is dened on an interval

J where J

J.
ii)

= for t J, and
iii)

satises the ordinary dierential equation on

J.
Theorem 2.3.1 Assume f : U R R
n
R
n
, U open and f(t, x) continuous and
locally Lipschitz with respect to the second variable. Then there exists a solution (t) of the
IVP
x
= f(t, x)
x(t
0
) = x
0
dened on an interval (m
1
, m
2
) with the property that if is any other solution of the IVP,
the domain of is contained in (m
1
, m
2
).
Proof Let M denote the set of all intervals on which solutions of the IVP are dened.
That M is not empty follows from the Existence Theorem. Let M
1
be the set of all right
hand endpoints of M and M
2
the set of all left hand endpoints. Take
m
1
= inf M
1
, m
2
= sup M
2
.
Pick any
t (m
1
, m
2
). Then there exists a solution of the IVP whose interval of denition
includes
t, say

. Dene a solution (t) on (m
1
, m
2
) by setting (
t) =
t). By uniqueness it
follows that (t) is well dened and is a solution for all t (m
1
, m
2
).
The interval (m
1
, m
2
) is called the maximal interval of existence corresponding to (t
0
, x
0
).
Furthermore, the maximal interval must be open (verify this).
Example 2.3.2 Take U to be the right half plane and consider
x
(t) =
1
t
2
cos(
1
t
)
x(t
0
) = x
0
Then x(t) = c sin(
1
t
) and the IVP can be solved for any inital condition (t
0
, x
0
), t
0
> 0.
Note that the maximal interval of existence is (0, ) and lim
t0
+ x(t) does not exist.
Example 2.3.3 Consider
x
= 3x
4/3
sin(t)
x(t
0
) = x
0
.
Solutions are x(t) 0 and x(t) = (c cos t)
3
where c is determined by the initial data
(t
0
, x
0
). Nontrivial solutions are dened on (, ) only if |c| > 1. Thus, the the maximal
interval of existence may depend on the initial conditions. Moreover, this example and
Example(2.3.1) suggest that the graph of a solution tends to innity at a nite endpoint of
the maximal interval of existence. This is indeed the case when f(t, x) is bounded, but the
complete story is a bit more involved. The next few theorems address this issue and clarify
these suggestions.
n
R
n
locally Lipschitz with respect to the second variable and bounded on U. If (t) is a solution
of the IVP,
x
(t) = f(t, x)
x(t
0
) = x
0
and dened for < t < , then the limits
lim
t
+
(t), lim
t
(t)
exist. If (, (
)), (, (
+
)) U, then the solution can be extended to the right and left.
Proof Let t
1
, t
2
(, ). Then
|(t
1
) (t
2
)| |
_
t
1
t
2
|f(s, (s))|ds|
B|t
1
t
2
|.
If we pick {t
n
} such that t
n

, then for any > 0,

|(t
n
) (t
m
)| B|t
n
t
m
| <
for all n, m suciently large. Hence {(t
n
)} is Cauchy and so converges. Thus lim
n
(t
n
)
exists. An identical argument applies for lim
t
(t).
The second assertion follows immediately from the remarks preceding the denition of
continuation.
Compare this theorem with the result of Example(2.3.2) in which f(t, x) =
1
t
2
cos(
1
t
)
was not bounded on U. As we observed, the solution did not have a limit at the left hand
endpoint of its maximal interval of existence.
n
R
n
locally Lipschitz with respect to the second variable and bounded on U. Let (m
1
, m
2
) denote
the maximal interval of existence of the solution of the IVP
x
= f(t, x)
x(t
0
) = x
0
.
Then either m
2
= or (m
2
, (m
2
)) is on the boundary of U. A similar statement holds for
m
1
.
Proof. First suppose m
2
< were nite. From the previous theorem, (m
2
) exists
and if (m
2
, (m
2
)) U then the solution could be extended to the right. It must follow that
(m
2
, (m
2
)) lies on the boundary of U. Similarly for m
1
.
Example 2.3.4 Reconsider the example
x
= x
2
x(0) = 1
Here U = R
2
and (t) =
1
1t
. Dene
U
A
= {(t, x) | |t| < , |x| < A}.
The maximal interval of existence is (m
1
, m
2
) = (, 1) and as t m
2
the graph of the
solution will always meet the boundary of U
A
when t = 1 1/A.
In general suppose f(t, x) is is continuous and locally Lipschitz with respect to the second
variable on all of R R
n
and the solution of an IVP has a maximal interval of existence,
(m
1
, m
2
) where m
2
< . One may modify the ideas in the previous example and apply
Theorem(2.3.2) to conclude that as t m
2
the graph of the solution always meets the
boundary |x| = A of the set U
A
. Since A can be arbitrarily large, the following theorem
must follow. (The details are left as an exercise.)
Corollary 2.3.1 Let U = RR
n
and (m
1
, m
2
) denote the maximal interval of existence
of the IVP. If |m
2
| < , then
lim
tm
2
|(t)| = .
(Similarly for m
1
).
This corollary provides a method for determining when a solution is global, that is,
dened for all time t. In particular, if f(t, x) is dened on all of R R
n
, then a solution is
global if it does not blow up in nite time. These ideas are illustrated in the next examples.
Example 2.3.5 Consider the equation for the damped, nonlinear pendulum.
y
(t) + y
+ sin y = 0, > 0
y(0) = y
0
, y
(0) = v
0
.
Rewrite the problem as a rst order system,
x
1
= y
x
2
= y
.
Then
x
=
d
dt
_
x
1
x
2
_
=
_
x
2
x
2
sin x
1
_
= f(x)
x(0) =
_
y
0
v
0
_
.
Since f
i
/x
j
are continuous for all (x
1
, x
2
), f is locally Lipschitz. Hence for any intitial
conditions the IVP has a unique solution. We now show the solution is global, i.e., it exists
for all t.
In a standard way, one rst multiplies the equation by y
to get
y
(y
+ y
+ sin y) = 0
and
d
dt
(
1
2
(y
)
2
cos y) = (y
)
2
0
or
d
dt
(
1
2
(y
)
2
+ (1 cos y)) 0.
Thus
1
2
(y
(t))
2
+ (1 cos y(t))
1
2
(y
(0))
2
+ (1 cos y(0))
=
1
2
(v
0
)
2
+ (1 cos y
0
).
Let
1 cos y
0
+
1
2
(v
0
)
2
=
1
2
p
2
0
and since (1 cos y) 0 we have,
1
2
(y
)
2
1
2
p
2
0
or
|y
| |p
0
|.
Since
y(t) = y
0
+
_
t
0
y
(s)ds
it follows that
|y(t)| |y
0
| +|t|p
0
and so |y(t)| < for all t.
Example 2.3.6 Consider the IVP
x
+ (x, x
)x
+ (x) = u(t)
x(0) = x
0
, x
(0) = v
0
where ,
x
,
x
, ,
are continuous and 0, z(z) 0. We will show that all solutions

are global.
First, it is a straightforward matter to verify that the IVP has a local solution for any
initial data. If we multiply the dierential equation by the solution, say (t), then
d
dt
(
1
2
(
)
2
+
_
(t)
0
(s) ds) = (,
)(
)
2
+ u(t)
(t)
u

1
2
(u
2
+ (
)
2
).
Since z(z) 0,
_

0
ds 0.
Call
F(t) =
1
2
(
)
2
+
_
(t)
0
(s) ds.
Then
F(t)
1
2
(
)
2
,
and from the above inequalities we see
F
(t)
1
2
((
)
2
+ u
2
) F(t) +
1
2
u
2
,
or
F
(t) F(t)
1
2
u
2
.
Thus
d
dt
(e
t
F)
1
2
e
t
u
2
or
F(t) F(0) e
t
_
t
0
e
s
u
2
(s) ds.
Thus we may write
1
2
(
)
2
F(t) G(t)
or
|
(t)| H(t)
where G(t), H(t) are functions that are nite for all t. With this bound on the derivative
we then get
|(t)| |x
0
| +|
_
t
0
|
(s)| ds < , for all t.

The preceding examples and Theorem(2.3.3) are special cases of the next result.
n
R
n
locally Lipschitz with respect to the second variable. Let (t) be the solution of the IVP
x
= f(t, x)
x(t
0
) = x
0
and (m
1
, m
2
) its maximal interval of existence. If m
2
< and E is any compact subset of
U, then there exists an > 0 such that (t, (t)) is not in E if t > (m
2
) (and similarly for m
1
.
Proof. Consider the closed set U
c
= R
n+1
U and let d(E, U
c
) = > 0. Now pick a
closed set E
U such that E E
and d(E, E
) < /2.
We will assume that (t, (t)) E for all t (m
1
, m
2
) and obtain a contradiction. To
this end, choose M such that |f(x, t)| M for all (t, x) E
and select r < /2. Pick any

(
t, x) E and let
K
r
= J
r
(
t) B
r
( x).
Note that if (t, x) K
r
, max{|t
t|, |x x|} r < /2 and K

r
E
. The IVP has a

unique solution that exists on an interval |t
t| < where < r, M < r, L < 1 and L is a

Lipshitz constant on the set E
. Moreover, the same M and L will work for any (
t, x) since
K
r
E
. Now select

t (m
2
, m
2
). Then (
t, (
t)) E so the IVP

x
= f(t, x)
x(
t) = (
t)
has a unique solution (t) that exists on |t
t| . Then
(t) =
_
(t), m
1
< t <

t
(t),

t t <

t +
is a continuation of (t) dened on (m
1
,
t + ). But
t + > m
2
+ > m
2
contradicting the maximality of (m
2
, m
2
).
2.4 Dependence on Data
In an initial value problem
x
= f(t, x)
x(t
0
) = x
0
2.4. DEPENDENCE ON DATA 19
one might regard t
0
, x
0
and f(t, x) as measured values or inputs in the formulations of a
physical model. Consequently it is important to know if small errors or changes in this data
would result in small changes in the solutions of IVP. That is, does the solution depend
continuously on (t
0
, x
0
) and f(t, x) in some sense.
Denote the solution the IVP by (t, t
0
, x
0
) where
(t
0
, t
0
, x
0
) = x
0
.
We will show that under reasonable assumptions on f, is continuous in the variables to
t
0
, x
0
and small changes in f result in small changes in . The following theorem is an
indespensible result in the study of dierential equations and is central to our results of this
section.
Theorem 2.4.1 [Gronwalls Inequality] Let f
1
(t), f
2
(t), p(t) be continuous on [a,b] and
p 0. If
f
1
(t) f
2
(t) +
_
t
a
p(s)f
1
(s) ds, t [a, b],
then
f
1
(t) f
2
(t) +
_
t
a
p(s)f
2
(s) exp[
_
t
s
p(u)du] ds.
Proof. Dene
g(t) =
_
t
a
p(s)f
1
(s)ds,
so
g
(t) = p(t)f
1
(t) p(t)(f
2
(t) +
_
t
a
p(s)f
1
(s) ds).
We then get
g
(t) p(t)g(t) p(t)f

2
(t),
d
dt
(g(t)e
_
t
a
p(u)du
) p(t)f
2
(t)e
_
t
a
p(u)du
,
g(t)e
_
t
a
p(u)du
_
t
a
p(s)f
2
(s)e
_
s
a
p(u)du
ds
and
g(t)
_
t
a
p(s)f
2
(s)e
_
t
s
p(u)du
ds.
Now f
1
(t) f
2
(t) + g(t) and so the result follows.
There are some special cases of Gronwalls inequality that should be noted.
(1). If p(x) = k and f
2
(x) = are constant, then Gronwall gives
f
1
(x) e
k(xa)
(2). If
f
1
(x) k
_
x
a
f
1
(t)dt, k 0
then f
1
(x) 0.
(3). Suppose |z
(x)| |z(x)| for a x b and z(a) = 0, then
_
x
a
z
(t)dt
_
x
a
|z
(t)|dt
_
x
1
|z(t)|dt
and so
|z(x)|
_
x
a
|z(t)| dt.
It follows by (2), that |z(x)| 0.
Theorem 2.4.2 Suppose (t), (t) satisfy
y
= f(t, y)
y(t
0
) = y
0
z
= g(t, z)
z(t
0
) = z
0
where f, g : U R R
n
R
n
, are continuous and locally Lipschitz with respect to the
second variable with Lipschitz constant K. If
|f(t, u) g(t, u)| , (t, u) U,
then
|(t) (t)| |y
0
z
0
|e
K|tt
0
|
+

K
(e
K|tt
0
|
1).
Proof. First assume t t
0
. Then
(t) (t) = y
0
z
0
+
_
t
t
0
f(s, (s)) g(s, (s) ds
= y
0
z
0
+
_
t
t
0
[f(s, (s)) f(s, (s)] + [f(s, (s)) g(s, (s))] ds.
Thus
|(t) (t)| |y
0
z
0
| + (t t
0
) + K
_
t
t
0
|(s) (s)| ds.
Now apply Gronwall with
f
1
= | |
f
2
= |y
0
z
0
| + (t t
0
)
p = k.
Then
|(t) (t)| (t t
0
) +|y
0
z
0
| + K
_
t
t
0
((s t
0
) +|y
0
z
0
|)e
K(ts)
ds
= (t t
0
) +|y
0
z
0
| + K{((s t
0
) +|y
0
z
0
|)
e
K(ts)
K
t
t
0
+
_
t
t
0
e
K(ts)
ds}
= (t t
0
) +|y
0
z
0
| + K{(
(t t
0
) + (y
0
z
0
)
K
+
1
K
|y
0
z
0
|e
K(tt
0
)
} + (
e
K(ts)
K
)
t
t
0
= |y
0
z
0
|e
K(tt
0
)
+

K
(e
K(tt
0
)
1).
If t < t
0
, a similar argument gives
|(t) (t)| |y
0
t
0
|e
k(t
0
t)
+

K
(e
k(t
0
t)
1)
and the result follows.
Example 2.4.1 Consider the initial value problems,
(1)
_
y
= f(t, y) = 1 + t
2
+ y
2
, Ricattis Equation
y(0) = y
0
(2)
_
z
= g(t, z) = 1 + z
2
z(0) = y
0
Of course problem (2) is easily solved. If we were to approximate the solution to (2) by that
of (1) on the set
U = {(t, u)||t| < 1/2, |u| < 1},
we would like to estimate the error. In the notation of Theorem(2.4.2)
|f(t, u) g(t, u)| = |t
2
| <
1
4
=
Also
|
f
u
| = |2u| 2, |
g
u
| = |2u| 2
and so we can take the common Lipschitz constant to be K = 2. Then
|y(t) z(t)|

k
(e
K|tt
0
|
1)
1
2
(e
2(1)
1) 0.2.
If, however, we were to restrict |t| < 1/4 then we get a much better approximation,
|y(t) z(t)|
1
32
(1.6487 1) . 0203
Exercises for Chapter 2
1. A solution y = (x) to
y
+ sin(x)y
+ (1 + cos(x))y = 0
is tangent to the x-axis at x = . Find (x).
2. Show that the initial value problem
y
=
1
1 + y
2
, y(0) = 1
has a unique solution that exists on the whole line.
3. Consider the initial value problem
y
(x) + F
(y) = 0, y(x
0
) = y
0
, y
(x
0
) = v
0
(a) If F C
2
(R), carefully explain why the Fundamental Existence and Uniqueness
theorem guarantees that this initial value problem has a unique solution for any point
(x
0
, y
0
) R
2
.
(b) Suppose that F(u) > 0, u R. Prove that the solution to the initial value
problem exists for all x R.
4. Consider the equation
y
(x) =
xy
1 + y
2
+ sin(x).
(a) Explain why for each (x
0
, y
0
) R
2
there is a solution of the dierential equation
that satises y(x
0
) = y
0
that is dened in some neighborhood of x
0
.
(b) Show that any solution of the dierential equation satises
|y(x)| k
1
e
k
2
x
2
for constants k
1
, k
2
.
(c) Prove that each solution of the dierential equation can be extended to all of R.
5. Consider
y
+ q(x)y = 0
y(x
0
) = y
0
, y
(x
0
) = v
0
where q C[a, b], x
0
[a, b].
(a) Carefully explain why this problem has a unique solution.
(b) Show that if a solution has a zero in [a, b] it must be simple.
6. Consider the equation
y
+ (1 + ap(x))y = 0
where a is a nonnegative constant and p(x) C(R), |p(x)| 1. Let D be the domain
D = {(x, y)| 0 x , 0 y 1} and let y = (x) denote the solution of the initial
value problem
y
+ (1 + ap(x))y = 0, y(0) = 0, y
(0) = 1.
Suppose we approximate the solution of the initial value problem by sin(x) on the
domain D. Estimate (x) sin(x) for 0 x .
7. Estimate the error in using the approximate solution y(x) = e
x
3
/6
0 x 1/2 for the initial value problem
y
(x) + xy(x) = 0
y(0) = 1, y
(0) = 0

Existence Theory and Properties of Solutions: TH (N) (N TH

Hochgeladen von

Dokumentinformationen

Originalbeschreibung:

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

Existence Theory and Properties of Solutions: TH (N) (N TH

Hochgeladen von

Copyright:

Verfügbare Formate

Chapter 2

Existence Theory and Properties of

(t) = f(t, x(t)) (2.1.1)

(t) = f(t, (t)).

(t) = f(t, x(t))

(t) = (t , t +) R and assume

(t) = f(t, x(t))

(t) = f(t, (t)), t J

(t) = f(t, (t))

is continuous. Thus by the Fundamental Theorem of Calculus

(t) = f(t, (t)).

t, x). By the previous proposition,

(t) = f(t, x(t))

)) U. Now consider the IVP

f(s, (s)) ds for t < +

(t) = f(t, y(t)), < t < +

, then for any > 0,

are continuous and 0, z(z) 0. We will show that all solutions

(s)| ds < , for all t.

and select r < /2. Pick any

t|, |x x|} r < /2 and K

. The IVP has a

t| < where < r, M < r, L < 1 and L is a

. Moreover, the same M and L will work for any (

t)) E so the IVP

(t) p(t)g(t) p(t)f

(x)| |z(x)| for a x b and z(a) = 0, then

Das könnte Ihnen auch gefallen