BOUNDEDNESS AND STABILITY OF GENERALIZED LIENARD

EQUATION

OGUNDARE, B. S.

Abstract. In this paper, we give a suitable Lyapunov function for global

asymptotic stability as well as the boundedness of solutions of a class of sec-
ond order nonlinear system. With the Lyapunov function, conditions on the
nonlinear terms to guarantee global stability and boundedness of the solution
and its derivative were given. Our results include and improve on the existing
results in the literature.

1. Introduction
In this paper, we are concerned with the global asymptotic stability and bound-
edness of solutions of the Lienard equation
ẍ + f (x)ẋ + g(x) = p(t, x, ẋ), (1.1)
where functions f, g, and p are continuous and depend (at most) only on the
arguments displayed explicitly. Here and elsewhere, all the solutions considered
and all the functions which appear are supposed real. The dots indicate differenti-
ation with respect to t. Lienard equations appears in a number of physical models
and it is important in describing fluid mechanical and nonlinear elastic mechanical
phenomena. A lot of mathematicians had shown interest in studying the dynamical
behaviour of all kinds of Lienard equations. (see for instances [3], [5], [6], [7], [10],
[12], [13], [14], [15], [16], [18], [19], [21] and [22] and references therein).
The notions of stability and boundedness of solutions are fundamental in the
theory and application of differential equations. In this way, both concepts lead to
the real world applications. It is well known that the study of second order non-
linear differential equations has attracted the interest of many researchers. Several
authors have contributed to the study of boundedness of solutions to Lienard equa-
tions (cf. [5], [6], [18], [19],and references therein) via the Lyapunov second method
since it is one of the most effective and widespread tool for studying the qualitative
behaviour of solutions to nonlinear differential equations. Results relative to stabil-
ity, boundedness, convergence and periodic solutions to equations of the form (1.1)
have been obtained. (see for instance [1]-[25]). With the aid of Lyapunov’s second
method, Burton [4], Fonda & Zanolin [6], Napoles [18] and Napoles & Repilado[19]
discussed the boundedness of solutions. In [5], the global asymptotic stability of
the equation (1.1) (i.e p(t; x, ẋ) = 0). In [18], the ultimate boundedness of solu-
tions of the equation (1.1) was considered for the autonomous case while Ogundare
and Okecha [20] discussed the boundedness, periodicity and stability of solutions
to certain class of the equation (1.1).

2000 Mathematics Subject Classification. 34A34, 34D20, 34D23, 34D99.

Key words and phrases. boundedness, global asymptotic stability, Lienard Equation, Lyapunov
function.
1
2

The motivation for the current paper comes from the works of Chunji Li [5],
Juan E. Napoles Valdes [18] and Ogundare & Okecha [20]. A suitable complete
Lyapunov function is used to give conditions on the nonlinear terms to guarantee
global asymptotic stability and boundedness of solutions of the equation (1.1). Our
results extend the results obtained by Chunji [5] for the global asymptotic stability
of the null solution when p ≡ 0 and boundedness and ultimate boundedness of
all solutions when p 6= 0 in the equation (1.1). Furthermore our results comple-
ment existing results on qualitative behaviour of solutions of second order nonlinear
differential equations.
This paper is organized as follows. Our basic assumptions are presented in
Section 2 alongside with the main results. Section 3 is devoted to some preliminary
results and in Section 4, the proofs of the main theorems are given.
An associated system to the equation (1.1) of interest to us is
ẋ = φ(y) − F (x)
(1.2)
ẏ = −g(x) + p(t, x, ẋ)
where Z t
F (x) = f (x)dx
0

2. Formulation of Results
Now let the functions f, g and p be continuous and the following conditions
hold:
(i)
g(x) − g(0)
= G0 ≤ α ∈ I0 = [0, aα](x 6= 0 and g(0) = 0);
x
(ii)
F (x) − F (0)
= F0 ≤ β ∈ I0 , (x 6= 0 and F (0) = K 6= 0);
x
(iii)
φ(y) − φ(0)
= Φ0 ≤ γ ∈ I0 , (y 6= 0 and φ(0) = 0);
y
where α, β, γ are all positive.
Next we state our main results.
Theorem 2.1. Suppose that conditions (i)-(iii) are satisfied with p(t, x, ẋ) ≡ 0,
then the trivial solution of the equation (1.1) is globally asymptotically stable.
Theorem 2.2. In addition to conditions (i)-(iii) being satisfied, suppose that
(iv)
p(t, x, ẋ) ≡ p(t) and |p(t)| ≤ M,
for all t ≤ 0, then there exists a constant σ, (0 < σ < ∞) depending only on the
constants α, β and γ such that every solution of (1.1) satisfies
 Z t 2
2 2 − 21 σt 1
στ
x (t) + ẋ (t) ≤ e A1 + A2 |p(τ )| e 2 dτ
t0
for all t ≥ t0 , where the constant A1 > 0, depends on α, β and γ as well as on
t0 , x(t0 ), ẋ(t0 ); and the constant A2 > 0 depends on α, β and γ.
 3

Theorem 2.3. Suppose the conditions of the Theorem 2.2 are satisfied with con-
dition (iv) replaced with
(v)
|p(t, x, ẋ)| ≤ (|x| + |y|)θ(t),
Z t
where θ(t) is a non negative and continuous function of t such that θ(s)ds ≤
0
M < ∞ is satisfied with a positive constant M . Then, there exists a constant K0
which depends on M, K1 , K2 and t0 such that every solution x(t) of the equation
(1.1) satisfies
|x(t)| ≤ K0 , |ẋ(t)| ≤ K0
for sufficiently large t.
Remark: We wish to remark here that while the Theorem 2.1 is on the global
asymptotic stability of the trivial solution, the Theorems 2.2 and 2.3 deal with the
boundedness and ultimate boundedness of the solutions respectively.
Notations: Throughout this paper K, K0 , K1 , . . . K14 will denote finite positive
constants whose magnitudes depend only on the functions F , g and p as well as
constants α, β, γ and δ but are independent of solutions of the equation (1.1).
Ki0 s are not necessarily the same for each time they occur, but each Ki , i = 1, 2...
retains its identity throughout.

3. Preliminary Results
We shall use as a tool to prove our main results a Lypunov function V (x, y)
defined by
δ δ
2V = (1 + α)x2 + [(1 + α) + a2 ]y 2 − 2δxy (3.1)
a aα
where α, β and δ are positive real numbers with α(1 + α) > a2
The following lemmas are needed in the proofs of Theorems 2.1, 2.2 and 2.3.
Lemma 3.1. Subject to the assumptions of Theorem 2.1 there exist positive con-
stants Ki = Ki (β, γ, δ), i = 1, 2 such that
K1 (x2 + y 2 ) ≤ V (x, y) ≤ K2 (x2 + y 2 ). (3.2)
Proof. First, it is clear from the equation (3.1) that V (0, 0) ≡ 0. Re-arranging
equation (3.1) we have,

δ δ aδ
2V = [α(1 + α) − a2 ]x2 + (1 + α)y 2 + (y − x)2 (3.3)
aα aα α
From the equation (3.3), we note that
δ δ
2V ≥ [α(1 + α) − a2 ]x2 + (1 + α)y 2 . (3.4)
aα aα
It follows from the above that there exists a constant K1 such that
V ≥ K1 (x2 + y 2 ), (3.5)
where
δ
× min α(1 + α) − a2 , (1 + α) .

K1 = (3.6)
aα
1
On the other hand, using the inequality −|xy| ≤ (x2 + y 2 ) in equation (3.1),
2
we have,
4

δ δ
(1 + α)x2 +
2V ≤ [(1 + α) + a2 ]y 2 + δ(x2 + y 2 ) (3.7)
a aα
Consequently, it is not difficult to establish from the inequality (3.7) that

V ≤ K2 (x2 + y 2 ) (3.8)
where
δ
K2 = × max {α[(1 + α) + a], (1 + α) + a(a + α)} . (3.9)
aα
At last, combination of inequalities (3.5) and (3.8) gives

K1 (x2 + y 2 ) ≤ V (x, y) ≤ K2 (x2 + y 2 ). (3.10)



Lemma 3.2. In addition to assumptions of Theorem 2.1, let the condition (vi)
of the Theorem 2.2 be satisfied also. Then there are positive constants Kj =
Kj (, β, γ, δ)(j = 3, 4) such that for any solution (x, y) of the system (1.2),
d
V̇ |(1.2) ≡ dt V |(1.2) (x, y) ≤ −K3 (x2 + y 2 ) + [p(t)K5 − F (0)K4 ](|x| + |y|).
(3.11)
Proof. Differentiate (3.1) along the solution path of the system (2.2) to obtain
δ δ
(1 + α)xẋ +
V̇ = ((1 + α) + a2 )y ẏ − δ(xẏ + y ẋ) (3.12)
a aα
Using the system (1.2) in the equation (3.12)
δ δ
V̇ = (1+α)x(φ(y)−F (x))+ ((1+α)+a2 )y(−g(x)+p(t))−δy(φ(y)−F (x))−δx(−g(x)+p(t))
a aα
(3.13)
δ δ
= −δ(1 + α)xF (x) − δxg(x) − δyφ(y) − (1 + α)xφ(y) − [(1 + α) + a2 ]yg(x)
 a aα
δ
+δyF (x) + p(t) [(1 + α) + a2 ]y − δx
aα
(3.14)
δ δ
= − {(1 + α)F (x) + ag(x)} x − δyφ(y) − (1 + α)xφ(y)
a a  (3.15)
δ 2 δ 2
− [(1 + α) + a ]yg(x) + δyF (x) + p(t) [(1 + α) + a ]y − δx
aα aα
on using the hypotheses on φ, f, g and φ, we have
 
δ 2 2 δ 2
V̇ ≤ − {(1 + α)β + aα} x − δγy + p(t) [(1 + α) + a ]y − δx
a   aα 
δ δ 2
−φ(0) δy + (1 + α)x − g(0) δx + [(1 + α) + a ]y (3.16)
 a  aα
δ
−F (0) (1 + α)x − δy
a
The preceding inequality implies that
δ
≤ − K3 (x2 + y 2 ) − F (0)K4 (|x| + |y|) + p(t)K5 (|x| + |y|) (3.17)
a
 5

which is also expressed as

δ
≤ − K3 (x2 + y 2 ) + [p(t)K5 − F (0)K4 ](|x| + |y|) (3.18)
a

where
 K3 = max [(1 + α)(1 + γ) + a2 ], aγ ; K4 = max {(1 + α), a} and K5 =
2
max (1 + α + a ), aα for some constants a, α, and γ. Hence,
V̇ ≤ −K3 (x2 + y 2 ) + K6 (|x| + |y|) (3.19)
where
K6 = [p(t)K5 − F (0)K4 ] (3.20)
Note: When p(t) ≡ 0,
V̇ ≤ −K3 (x2 + y 2 ) − K4 F (0)(|x| + |y|) (3.21)
V̇ ≤ −K3 (x2 + y 2 ) (3.22)
Hence the proof of lemma 3.2. 

4. Proof of Main Results

We shall now give the proofs of the main results.
Proof of Theorem 2.1. The proof of Theorem 2.1 follows from Lemmas 3.1 and
3.2 where it has been established that the trivial solution of the equation (1.1) is
globally asymptotically stable. i.e every solution (x(t), ẋ(t)) of the system (1.2)
satisfies x2 (t) + ẋ2 (t) −→ 0 as t −→ ∞. 
Proof of Theorem 2.2. Indeed, by using the inequality (3.19), it follows that
dV 1
≤ −K3 (x2 + y 2 ) + K7 (x2 + y 2 ) 2 |p| .
dt
 
K4 F (0)
where K7 = 1 − since p(t) 6= 0.
K5 p(t)
Again, it also follows from the inequality (3.8) that
 1
2 2 21 V 2
(x + y ) ≤ .
K1
Thus, the inequality (3.19) becomes
1
V̇ ≤ −K8 V + K9 V 2 |p| . (4.1)
V
We note that K3 (x2 + y 2 ) = K3 · K1 and
dV 1
≤ −K8 V + K9 V 2 |p| , (4.2)
dt
K3 K7
where K8 = K1 and K9 = 1 .
K22
Furthermore, the above inequality can be written as
1
V̇ ≤ −2K10 V + K9 V 2 |p| , (4.3)
where K10 = 12 K8 .
Therefore
1
V̇ + K10 V ≤ −K10 V + K9 V 2 |p| . (4.4)
6

K10
Choosing a constant K11 such that K11 = K9 gives
1
n 1
o
V̇ + K10 V ≤ K9 V 2 |p| − K11 V 2 . (4.5)

Thus the inequality (4.5) becomes

1
V̇ + K10 V ≤ K9 V 2 V ∗ (4.6)
where
1
V ∗ = |p| − K11 V 2 (4.7)
1
≤V 2 |p|
≤ |p| . (4.8)
1
When |p| ≤ K11 V , 2

V ∗ ≤ 0, (4.9)
1
and when |p| ≥ K11 V , 2

1
V ∗ ≤ |p| ·
. (4.10)
K11
On substituting the inequality (4.9) into the inequality (4.5), we have
1
V̇ + K10 V ≤ K12 V 2 |p| ,
where
K9
K12 = .
K11
This implies that
1 1
V − 2 V̇ + K10 V 2 ≤ K12 |p| . (4.11)
1
Multiplying both sides of the inequality (4.11) by e , gives 2 K10 t

1
n 1 1
o 1
e 2 K10 t V − 2 V̇ + K10 V 2 ≤ e 2 K10 t K12 |p| , (4.12)
i.e
d n 1 1 K10 t o 1
2 V 2 e2 ≤ e 2 K10 t K12 |p| . (4.13)
dt
Integrating both sides of (??) from t0 to t, gives
ot Z t
n 1 1
K10 γ 1 1 K10 τ
V e2 2 ≤ e2 K12 |p(τ )dτ | , (4.14)
t0 t0 2
which implies that
Z t
n 1 o 1 1 1 1 1
V 2 (t) e 2 K10 t ≤ V 2 (t0 )e 2 K10 t0 + K12 |p(τ )| e 2 K10 τ dτ ,
2 t0
or  Z t 
1 1 1 1 1 1
V 2 (t) ≤ e− 2 K10 t V 2 (t0 )e 2 K10 t0 + K12 |p(τ )| e 2 K10 τ dτ .
2 t0
On utilizing inequalities (3.8) and (3.10), we have
K1 (x2 (t) + ẋ2 (t)) ≤
1
e− 2 K10 t K2 (x2 (t0 ) + ẋ2 (t0 ))

Z t 2 (4.15)
1 1
K10 t0 1 K10 τ
e 2 + 2 K12 |p(τ )| e 2 dτ ,
t0
 7

for all t ≥ t0 .
Thus,
x2 (t) + ẋ2 (t) ≤ n n
1 − 21 K10 t 1
K1 e K2 (x2 (t0 ) + ẋ2 (t0 ))e 2 K10 t0
Z t 2
1
1 K τ
+ 2 K12 |p(τ )| e 2 10
dτ }
t0
(  Z t 2 )
− 12 K10 t 1
2 K10 τ
≤ e A1 + A2 |p(τ )| e dτ , (4.16)
t0

where A1 and A2 are constants depending on {K1 , K2 , . . . K12 and (x2 (t0 )+ ẋ2 (t0 )).
By substituting K10 = µ in the inequality (4.16), we have
(  Z t 2 )
2 2 − 21 µt 1
µτ
x (t) + ẋ (t) ≤ e A1 + A2 |p(τ )| e 2 dτ .
t0

This completes the proof. 

Proof of Theorem 2.3. From the definition of function V and the conditions of The-
orem 2.3, the conclusion of Lemma 3.1 can be obtained as
V ≥ K1 x2 + y 2 ,


(4.17)
and since P 6= 0 we can revise the conclusion of Lemma 3.2, i.e,
V̇ ≤ −K3 (x2 + y 2 ) + K7 (|x| + |y|) |p| ,
and we obtain by using the condition (v) as stated in the Theorem 2.3 that
V̇ ≤ K7 (|x| + |y|)2 θ(t). (4.18)
1 2
Using the inequality |xy| ≤ (x + y 2 ) on (4.18), we have
2
V̇ ≤ K13 (x2 + y 2 )θ(t), (4.19)
where K13 = 2K7 .
From inequalities (4.17) and (4.19) we have,
V̇ ≤ K13 V θ(t). (4.20)
Integrating inequality (4.20) from 0 to t gives
Z t
V (t) − V (0) ≤ K14 V (s)θ(s)ds, (4.21)
0
K13 2K7
where K14 = K1 = K1 . Thus,
Z t
V (t) ≤ V (0) + K14 V (s)θ(s)ds. (4.22)
0

Applying the Grownwall-Reid-Bellman theorem on the inequality (4.22) yields

 Z t 
V (t) ≤ V (0)exp K14 θ(s)ds . (4.23)
0

This completes the proof of Theorem 2.3. 

8

References
[1] F.A. Atkinson: On Second Order Non-linear Oscillations, Pacific J. Math. 5(1955), 643-647.
[2] R. J. Ballieu and K. Peiffer : Attractivity of the Origin for the Equation x00 +
f (t, x, x0 )|x0 |α x0 + g(x) = 0 , J. Math. Anal. Appl. 65(1978), 321-332.
[3] I. Bihari: On Periodic Solutions of Certain Second Order Ordinary Differential Equations
with Periodic Coefficients, Acta Math. Acad. Sci. Hung., 11(1960), 11-16.
[4] T. A. Burton : Liapunov functions and Boundedness, J. Math. Anal. Appl. 58(1977), 88-97.
[5] L. Chunji: Global Asymptotic Stability of Generalized Lienard Equation,Trends in Mathe-
matics, Information Center for Mathematical Sciences, Volume 4, Number 2, December 2001,
127-131.
[6] A. Fonda and F. Zanolin: Bounded Solutions of Nonlinear Second Order Ordinary Differen-
tial Equation, Discrete and Continuous Dynamical Systems, Vol. 4, No. 1 1998, 91-98.
[7] Guo Rong Ye, Hui-Sheng Ding, Xi-Lang Wu: Uniform Boundedness of Solutions for a Class
of Lienard Equations, Electronic Journal of Differential Equations, Vol. 2009 (2009), No. 97,
1-5.
[8] J. R. Haddock : On Liapunov functions for Non autonomous Systems,J. Math. Anal. Appl.
47(1974), 599-603.
[9] W. Hahn: Theory and Application of Liapunov’s Direct Method, New Jersey, Prentice-Hall,
Inc., 1963.
[10] L. Hatvani: On the Stability of the zero Solutions of Certain Second Order differential Equa-
tions, Acta Sci. Math., 32(1971), 1-9.
[11] L. Hatvani: On  the Strong Stability by Lyapunov’s direct method, Acta Scientiarum Math-
ematicarum, 37(1975), 241-251.
[12] L. Hatvani: On  Stability Properties of Solutions of Second Order Differential Equations,
EJQDE, Proc. 6th Coll. QTDE, 11(2000), 1-6.
[13] L. Hatvani, T. Krisztin and V. Totik: A necessary and sufficient condition for the Asymptotic
Stability of the Damped Oscillator , J.Differential Equations (2)119 (1995), 209-223.
[14] J. Karsai: On the global Asymptotic of the zero Solution of the Equation x00 + g(t, x, x0 )x0 +
f (x) = 0, Studia Sci. Math. Hungar. 19(1984), 385-393.
[15] J. Karsai: On the Asymptotic Stability of the zero Solution of Certain Nonlinear Second Or-
der Differential Equation , Differential Equations: Qualitative Theory (Szeged, 1984),Colloq.
Math. Soc. Janos Bolyai, 47, Amsterdam and New York, North-Holland, (1987), 495-503.
[16] A. Kroopnick: Properties of Solutions to a Generalized Lienard Equation with Forcing Term,
Applied Mathematics E-Notes, 8(2008), 40-44.
[17] Gh. Morosanu, C. Vladimirescu: Stability for a Nonlinear Second Order ODE, Funkcialaj
Ekvacioj, 48(2005), 49-56.
[18] J. E. Napoles Valdes: On the ultimate Boundedness of Solutions of systems of Differential
Equations, Revista INTEGRATION Universidad Industrial de Santander Escula de Mathe-
maticas, Vol. 13, No. 1, 41-47, ener-junio de 1995.
[19] J. E. Napoles and J. A. Repilado : On the Boundedness and Asymptotic Stability in the
Whole of the Solutions of a System of Differential Equations, Rev. Ciencias Matematicas,
Univ. of Habana, Vol. XVI, No.1(1995), 83-86.
[20] B. S. Ogundare and G. E. Okecha: Boundedness, Periodicity and Stability of Solutions to
ẍ + a(t)g(ẋ) + b(t)h(x) = p(t; x, ẋ), Math. Sci. Res. J. 11(5) 2007, 432-443.
[21] P. Pucci and J. Serrin: Precise Damping conditions for global Asymptotic Stability for Non-
linear Second Order Systems, Acta Mathematica, 170(1993), 275-307.
[22] P. Pucci and J. Serrin: Precise Damping conditions for global Asymptotic Stability for Non-
linear Second Order Systems II, Journal of Differential Equations, 113(1994), 815-835.
[23] R. Smith: Asymptotic Stability of x00 + a(t)x0 + x = 0, Quarterly Journal of Mathematics,
Oxford Series, (2) 12(1961), 123-126.

[24] C. Vladimirescu: Stability for Damped Oscillators, Annals of University of Craiova, Math.
Comp. Sci. Ser. Vol. 32(2005), 227-232.
[25] P. Waltman: Some Properties of Solutions of u00 + a(t)f (u) = 0, Monatshefte Fur Matem-
atiK, 67(1963), 50-54.
[26] J. S. W. Wong: A note on Boundedness theorem of Certain Second Order Differential Equa-
tions, Siam Review, 6(1964), 175-176.
 9

[27] J. S. W. Wong and T. A. Burton: Some Properties of Solutions of u00 + a(t)f (u)g(u0 ) = 0,
Monatshefte Fur Mathematik, 69(1965), 368-374.
[28] T. Yoshizawa : Stability Theory by Liapunov’s Second Method The Mathematical Society of
Japan (1966).

Ogundare, B. S.
Department of Mathematics, Obafemi Awolowo University, Ile-Ife, Nigeria
E-mail address: [email protected], [email protected]