See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/271935531

Uniqueness of solutions for second order differential equations

Article in Monatshefte für Mathematik · July 2014

DOI: 10.1007/s00605-014-0661-z

CITATIONS READS

4 1,463

2 authors:

Daniel C. Biles John Spraker

Belmont University Western Kentucky University
36 PUBLICATIONS 561 CITATIONS 22 PUBLICATIONS 106 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by John Spraker on 27 February 2015.

The user has requested enhancement of the downloaded file.

Monatsh Math
DOI 10.1007/s00605-014-0661-z

Uniqueness of solutions for second order differential

equations

Daniel C. Biles · John S. Spraker

Received: 3 May 2014 / Accepted: 7 June 2014

© Springer-Verlag Wien 2014

Abstract We present a theorem concerning uniqueness of solutions for second order

ordinary differential equations with initial conditions. Despite the fact that the
hypotheses are weak and the theorem covers a large number of cases, the proof is
not especially complicated.

Keywords Uniqueness · Differential equations · Second order · Initial value

problem · Lipschitz

Mathematics Subject Classification (2010) 34A12 · 34A34 · 34A36

We consider uniqueness of solutions for the initial value problem

x  (t) = f (t, x(t), x  (t)), t ∈ [0, T ] (IVP)

x(0) = 0
x  (0) = ψ

where f : [0, 1] × R × R → R and ψ ∈ R. We define a solution as follows:

Communicated by A. Constantin.

D. C. Biles (B)
Department of Mathematics and CS, Belmont University, 1900 Belmont Blvd,
Nashville, TN 37212, USA
e-mail: [email protected]

J. S. Spraker
Mathematics Department, Western Kentucky University, 1906 College Heights Blvd,
Bowling Green, KY 42101, USA
e-mail: [email protected]

123
 D. C. Biles, J. S. Spraker

Definition x : [0, 1] → R is a solution of (IVP) on [0, T ], where T > 0 if

(i) x ∈ C 1 ([0, T ], R) and x  ∈ AC ([0, T ], R),
(ii) x  (t) = f (t, x(t), x  (t)) a.e. on [0, T ],
(iii) x(0) = 0 and x  (0) = ψ.

The monograph [1] gives an excellent survey of the uniqueness results for ordinary
differential equations up to its time of publication (1993), and includes well-known
results by Bernfeld/Driver/Lakshmikantham, Brauer, Kamke, Krasnosel’skii/Krein,
Montel/Tonelli, Nagumo, Osgood, Peano, Perron, Rogers, Van Kampen, Wend, Wint-
ner and many others. More recently, uniqueness results have been proven in [2] which
generalized Wintner’s result for nth order systems, [3] which furthered generaliza-
tions of Nagumo’s classical theorem, [4] for nth order systems, [5] for a fourth order
problem, [6] where Osgood’s condition was considered for uniqueness of bounded
-variation solutions, [7] where the initial value problem for first order systems was
studied, [8] for nth order systems under a Lipschitz type assumption and [9] where
second order problems were further studied. [3] was extended in [10]. A recent result
for fractional order differential equations was proven by [11].
This article was originally motivated by [12].
The proof of the following lemma follows that found in [1].

Lemma Let g : [0, ∞) → [0, ∞) be continuous and nondecreasing with g(0) = 0

and g(x) > 0 for x > 0 which satisfies

β
du
=∞ (1)
g(u)
0

for any β > 0. Let a > 0 and let y : [0, a] → [0, ∞) be continuous and assume that
there is an integrable function B : [0, a] → [0, ∞) such that
 t
y (t) ≤ B(τ )g (y(τ )) dτ
0

for all t ∈ [0, a]. Then, y ≡ 0 on [0, a].

We now prove the following Osgood-type uniqueness theorem.

Theorem 1 Assume ψ ≥ 0. Let f : [0, 1] × R × R → R, m 1 > 0, m 2 > ψ be given

such that
(1) f (t, x, y) > −ψ for almost all t ∈ [0, 1], all x ∈ R with 0 ≤ x ≤ m 1 and all
y ∈ R with 0 < y ≤ m 2 ,
(2) | f (t, x1 , y1 ) − f (t, x2 , y2 )| ≤ A (t, x1 , x2 , y1 , y2 ) g (|x1 − x2 | + |y1 − y2 |),
for almost all t ∈ [0, 1], all x1 , x2 ∈ ((ψ/2) t, m 1 ] and all y1 , y2 ∈ (0, m 2 ],

123
Uniqueness of solutions for second order differential equations

where
(a) A : [0, 1] × (0, m 1 ] × (0, m 1 ] × (0, m 2 ] × (0, m 2 ] → [0, ∞) satisfies
A (·, h 1 (·), h 2 (·), h 3 (·) , h 4 (·)) is integrable on [0, 1], where
h 1 , h 2 ∈ C 1 ([0, 1], ((ψ/2) t, m 1 ]) and h 3 , h 4 ∈ AC ([0, 1], (ψ − ψt, m 2 ]),
and
(b) g : [0, ∞) → [0, ∞) is continuous and nondecreasing with g(0) = 0 and
g(x) > 0 for x > 0 which satisfies

β
dx
=∞
g(x)
0

for any β > 0.

Then, there exists a T ∈ (0, 1] such that (IVP) has at most one solution on [0, T ].
 
Proof Let x1 , x2 be solutions of (IVP) with a common interval of existence 0, t¯
with 0 < t¯ ≤ 1.
 
Case 1 ψ > 0 : Choose t1 ∈ 0, t¯ so that x1 (t), x2 (t) ∈ (0, m 1 ] and x1 (t), x2 (t) ∈
(0, m 2 ] for all t ∈ (0, t1 ] (recall the initial conditions).
Case 2 ψ = 0 : By assumption (1), we have x1 (t) > 0 for almost all t ∈ (0, t¯], hence
x1 (t) > 0 and x1 (t) > 0 for t ∈ (0, t¯], and similarly for x2 . Choose
t1 ∈ 0, t¯ so that x1 (t), x2 (t) ∈ (0, m 1 ] and x1 (t), x2 (t) ∈ (0, m 2 ] for
all t ∈ (0, t1 ].
So, in either case we have a t1 such that x1 (t), x2 (t) ∈ (0, m 1 ] and x1 (t), x2 (t) ∈
(0, m 2 ] for all t ∈ (0, t1 ]. Define x : [0, t1 ] → R by x(t) = x1 (t) − x2 (t). For all
t ∈ [0, t1 ] and w equal to either x1 or x2 , we have

t

 
w (t) = ψ + f τ, w(τ ), w  (τ ) dτ . (2)
0
 
(Note that by our definition of solution, w ∈ AC [0, t¯], R , hence w  (·) =
f (·, w(·), w  (·)) exists a.e. and is integrable on [0, t].) Hence w  (t) > ψ − ψt,
and

 t s
 
w(t) = ψt + f τ, w(τ ), w  (τ ) dτ ds
0 0
t t t
   
= ψt + f τ, w(τ ), w  (τ ) ds dτ =ψt + (t−τ ) f τ, w(τ ), w (τ ) dτ
0 τ 0
t  
t2 ψt
> ψt − (t − τ )ψ dτ = ψ t − ≥ . (3)
2 2
0

123
 D. C. Biles, J. S. Spraker

From (2) and (3), we then have

|x(t)| + x  (t)
t
≤ (t − τ ) f (τ, x1 (τ ), x1 (τ )) − f (τ, x2 (τ ), x2 (τ )) dτ
0
t
+ f (τ, x1 (τ ), x1 (τ )) − f (τ, x2 (τ ), x2 (τ )) dτ
0
t
   
≤ (t − τ + 1)A τ, x1 (τ ), x2 (τ ), x1 (τ ), x2 (τ ) g |x(τ )| + x  (τ ) dτ
0
t
 
≤ 2 Ã(τ ) g |x(τ )| + x  (τ ) dτ
0

 
where Ã(τ ) = A τ, x1 (τ ), x2 (τ ), x1 (τ ), x2 (τ ) is integrable. Now, letting y(·) =
|x(·)| + x  (·) , it follows that

t
y(t) ≤ 2 Ã(τ ) g (y(τ )) dτ.
0

Applying the Lemma with a = t1 and B(τ ) = 2 Ã(τ ), we conclude y ≡ 0 on [0, t1 ],

i.e., x1 ≡ x2 on [0, t1 ].

Notes:
1. Theorem 1 generalizes the one in [12]. To verify that the Lipschitz-type condition
√ is a special case of assumption (2a), choose A (t,1x1 , x2 ) ≡ 1/(t + 1) +
in [12]
C/ min {x1 , x2 } (where C > 0) and then for√ h 1 , h 2 ∈ C ([0, 1], ((ψ/2)t, m 1 ]),
we √have A(t, h 1 (t), h 2 (t)) = 1/(t + 1) + C/ min {h 1 (t), h 2 (t)} ≤ 1/(t + 1) +
C/ (ψ/2)t, which is integrable on [0, 1].
2. A similar theorem can be proven for the case ψ < 0.
3. For ψ = 0, the reader might be tempted to use f (t, x, y) ≥ 0 in assumption (1)
so that it reads

f (t, x, y) ≥ 0 for almost all t ∈ [0, 1], all x ∈ R with 0 ≤ x ≤ m 1 and

all y ∈ R with 0 < y ≤ m 2 .

2, for x > 0
However, if we let f (t, x, y) = f (x) = , we get solutions x ≡ 0 and
0, for x ≤ 0
x = t 2 . Note that this example also shows that we cannot remove x = 0 in assumption

123
Uniqueness of solutions for second order differential equations

(1) so that it reads

f (t, x, y) > 0 for almost all t ∈ [0, 1], all x ∈ R with 0 < x ≤ m 1 and
all y ∈ R with 0 < y ≤ m 2 .

Acknowledgments The authors would like to thank the referee for his or her helpful suggestions.

References

1. Agarwal, R.P., Lakshmikantham, V.: Uniqueness and Nonuniqueness Criteria for Ordinary Differential
Equations. World Scientific, New Jersey (1993)
2. Constantin, Adrian: On the unicity of solutions for the differential equation x (n) (t) = f (t, x). Rendi-
conti del Circolo Matematico di Palermo Serie II 42, 59–64 (1993)
3. Constantin, A.: On Nagumo’s theorem. Proc. Jpn. Acad. 86, 41–44 (2010)
4. Eloe, P., Henderson, J., Khan, R.A.: Existence and uniqueness conditions for a class of (k+4j)-point
n-th order boundary value problem. Nonlinear Dyn. Syst. Theory 12, 49–62 (2012)
5. Gao, Y., Gu, Y.: Existence and uniqueness theorems of solutions for fourth-order boundary value
problems. Adv. Differ. Equ. Control Process. 5, 101–113 (2010)
6. Liang, X., He, W., Li, B.: Uniqueness of bounded -variation solutions for a class of discontinuous
system. Gongcheng Shuxue Xuebao 29, 757–762 (2012)
7. Liu, L.S.: Existence and uniqueness of solutions of discontinuous first-order differential system. Math.
Appl. (Wuhan) 8, 273–277 (1995)
8. Pan, Y., Wang, M.: Uniqueness for n-th order differential systems with strong singularities. Electr. J.
Differ. Equ. 2010, 1–9 (2010)
9. Pražák, D.: Remarks on the uniqueness of second order ODE. Appl. Math. 56, 161–172 (2011)
10. Mustafa, O.G., O’Regan, D.: On the Nagumo uniqueness theorem. Nonlinear Anal. 74, 6383–6386
(2011)
11. Băleanu, D., Mustafa, O.G., O’Regan, D.: A uniqueness criterion for fractional differential equations
with Caputo derivative. Nonlinear Dyn. 71, 635–640 (2013)
12. Mustafa, O.G.: On the uniqueness of flow in a recent tsunami model. Appl. Anal. 91, 1375–1378
(2012)

123
View publication stats