We present a survey on the existence of nontrivial solutions to impulsive differential equations by using variational methods, including solutions to boundary value problems, periodic solutions, and homoclinic solutions.
1. Introduction
There are many processes and phenomena in the real world, which are subjected during their development to the short-term external influences. Their duration is negligible compared with the total duration of the studied phenomena and processes. Therefore, it can be assumed that these external effects are “instantaneous”; that is, they are in the form of impulses. The investigation of such “leaps and bounds” developing dynamical states is a subject of different sciences: mechanics, control theory, pharmacokinetics, epidemiology, population dynamics, economics, ecology, and so forth. We refer the reader to [1–6].
During the last two decades, the theory of the existence of solutions for impulsive differential equations has attracted much attention because of its important applications; see [3, 7–12]. Classical approaches to such problems include fixed point theory and the method of upper and lower solutions. In 2008, Tian and Ge first applied variational methods in the study of boundary value problems (BVPs) of impulsive differential equations. From then on, variational methods have been widely used to study impulsive problems, such as boundary value problems, periodic solutions, and homoclinic solutions. We refer the reader to [3, 13–22].
The main aim of this survey is to present some recent existence results for impulsive differential equations via variational methods. More precisely, we will explore the variational framework of impulsive differential equations and study the existence and multiplicity of solutions of boundary value problems, periodic solutions, and homoclinic solutions.
The rest of this paper is organized as follows. In Section 2, we will state some important results for the second-order impulsive differential equations with two kinds of boundary conditions. Periodic solutions to impulsive problems will be considered in Section 3. In Section 4, homoclinic solutions to impulsive differential equations will be studied.
Finally in this section, we must note that besides the results presented in this survey many interesting and important results on impulsive differential equations via variational methods have been obtained by other researchers; see, for example, [23–31] and the references cited therein.
2. Boundary Value Problems
In this section, we recall some results of boundary value problems for impulsive differential equations.
Firstly, we consider a class of impulsive problems with Dirichlet boundary condition
(1)-u′′(t)+λu(t)=f(t,u(t))+g(t),a.e.t∈[0,T],Δu′(tj)=u′(tj+)-u′(tj-)=Ij(u(tj)),j=1,2,…,p,u(0)=u(T)=0,
where λ is a constant, f:[0,T]×R→R is continuous, g∈L2[0,T] and Ij:R→R, j=1,2,...,p, are continuous.
From now on, we assume that λ>-(π2/T2) and define
(2)L(u,u)=12∫0T|u′(t)|2dt+λ2∫0T|u(t)|2dt,000000000000000000foranyu∈H01(0,T).
Then we have the following.
Proposition 1 (see [21, Proposition 1.1]).
If λ>-(π2/T2), then there exist constants b>a>0 such that
(3)a∥u∥H012≤L(u,u)≤b∥u∥H012,foranyu∈H01(0,T).
Definition 2.
One says that a function u∈H01(0,T) is a weak solution to problem (1) if
(4)∫0T(u′(t),v′(t))+λ(u(t),v(t))dt+∑j=1pIj(u(tj))v(tj)=∫0Tf(t,u)vdt+∫0Tg(t)vdt
holds for any v∈H01(0,T).
Theorem 3 (see [15, Theorem 4.2]).
Suppose that g≡0, f is bounded, and the impulsive functions Ij are bounded. Then problem (1) has at least one weak solution to λ>-(π2/T2).
If f is sublinear at infinity on u and the impulsive functions Ij have sublinear growth, then we have the following.
Theorem 4 (see [15, Theorem 4.3]).
Assume that g≡0 and the following conditions are satisfied.
There exist a,b>0 and γ∈(0,1) such that
(5)|f(t,u)|≤a+b|u|γ,forevery(t,u)∈[0,T]×R;
there exist aj,bj>0 and γj∈[0,1)(j=1,2,…,p) such that
(6)|Ij(u)|≤aj+bj|u|γj,foreveryu∈R(j=1,2,…,p).
Then problem (1) has at least one weak solution to λ>-(π2/T2).
If f is superlinear at infinity on u, then we have the following.
Theorem 5 (see [21, Theorem 1.2]).
Assume that (I1) holds and the following conditions are satisfied.
There is a constant μ>2 such that 0<μF(t,u)≤uf(t,u) for every t∈[0,T] and u∈R∖{0};
g∈L2[0,T] satisfies the following inequality
(7)a-Mλ1-∑j=1pT(aj+bj)-T∥g∥L2λ1>0,
where M≔sup{F(t,u):t∈[0,T],|u|=1} and a is defined in Proposition 1.
Then problem (1) has at least one weak solution to λ>-(π2/T2).
Furthermore, we have the following result of multiple solutions.
Theorem 6 (see [21, Theorem 1.3]).
Let g(t)≡0. Assume that the conditions (F2), (I1), and (G) hold. Moreover, f and the impulsive functions Ij are odd about u; then problem (1) has infinitely many solutions.
Next, we consider another Dirichlet impulsive problem
(8)-u′′(t)+q(t)u(t)=f(t,u(t)),a.e.t∈[0,T],Δu′(tj)=u′(tj+)-u′(tj-)=Ij(u(tj)),j=1,2,…,p,u(0)=u(T)=0,
where q∈L∞[0,T], f:[0,T]×R→R is continuous, and Ij:R→R, j=1,2,…,p, are continuous.
Theorem 7 (see [20, Theorem 3.1]).
Let dj(j=1,2…,p) be fixed constants. If f(t,u)=σ(t)∈L2(0,T) and Ij(u)=dj(j=1,2,…,p); then problem (8) has a unique solution u if ess inft∈[0,T]q(t)>-(π2/T2).
Similar to Theorem 4, we have the following.
Theorem 8 (see [20, Theorem 3.2]).
Assume that conditions (F1) and (I1) hold. Then problem (8) has at least one weak solution to ess inft∈[0,T]q(t)>-(π2/T2).
If condition (I1) is replaced by the following condition:
Ij(j=1,2…,p) satisfy ∫0uIj(s)ds≥0 for all u∈R,
then we have the following.
Theorem 9 (see [30, Theorem 3.1]).
Assume that conditions (F1) and (I3) hold. Then problem (8) has at least one weak solution to ess inft∈[0,T]q(t)>-(π2/T2).
Now we study a class of impulsive Hamiltonian systems with periodic boundary conditions
(9)-u¨+A(t)u=λ∇F(t,u)+μ∇G(t,u),a.e.t∈[0,T],Δ(u˙i(tj))=u˙i(tj+)-u˙i(tj-)=Iij(ui(tj)),000000000000i=1,2,…,N,j=1,2,…,l,u(0)-u(T)=u˙(0)-u˙(T)=0,
where A:[0,T]→RN×N is a continuous map from the interval [0,T] to the set of N-order symmetric matrices, λ, μ are two positive parameters, T is a real positive number, u(t)=(u1(t),u2(t),…,uN(t)), tj,j=1,2,…,l, are the instants where the impulses occur and 0=t0<t1<t2<⋯<tl<tl+1=T, Iij:R→R(i=1,2,…,N,j=1,2,…,l) are continuous, and F,G:[0,T]×RN→R are measurable with respect to t, for every u∈RN, continuously differentiable in u, for almost every t∈[0,T], and satisfy the following standard summability condition:
(10)sup|u|≤bmax{|F(·,u)|,|G(·,u)|,|∇F(·,u)|,000000000|∇G(·,u)|}∈L1(0,T),∀b>0.
If X is a real Banach space, denote by WX the class of all functionals Φ:X→R possessing the following property: if {un} is a sequence in X converging weakly to u∈X and liminfn→∞Φ(un)≤Φ(u), then {un} has a subsequence converging strongly to u.
For example, if X is uniformly convex and g:[0,+∞)→R is a continuous, strictly increasing function, then, by a classical result, the functional u→g(∥u∥) belongs to the class WX.
Theorem 10 (see [32]).
Let X be a separable and reflexive real Banach space; let Φ:X→R be a coercive, sequentially weakly lower semicontinuous C1 functional, belonging to WX, bounded on each bounded subset of X, whose derivative admits a continuous inverse on X*; J:X→R is a C1 functional with compact derivative. Assume that Φ has a strict local minimum u0 with Φ(u0)=J(u0)=0. Finally, setting
(11)α=max{0,limsup∥u∥→+∞J(u)Φ(u),limsupu→u0J(u)Φ(u)},β=supu∈Φ-1((0,+∞))J(u)Φ(u),
assume that α<β.
Then, for each compact interval [a,b]⊂(1/β,1/α) (with the conventions 1/0=+∞,1/+∞=0), there exists R>0 with the following property: for every λ∈[a,b] and every C1 functional Ψ:X→R with compact derivative, there exists δ>0 such that, for each μ∈[0,δ],
(12)Φ′(x)=λJ′(x)+μΨ′(x)
has at least three solutions in X whose norms are less than R.
The following two results of Ricceri guarantee the existence of three solutions for a given equation.
Theorem 11 (see [33]).
Let X be a reflexive real Banach space; I⊆R is an interval; let Φ:X→R be a sequentially weakly lower semicontinuous C1 functional, bounded on each bounded subset of X, whose derivative admits a continuous inverse on X*; -J:X→R is a C1 functional with compact derivative. Assume that
(13)lim∥x∥→+∞(Φ(x)-λJ(x))=+∞
for all λ∈I and that there exists ρ∈R such that
(14)supλ∈Iinfx∈X(Φ(x)+λ(ρ-J(x)))<infx∈Xsupλ∈I(Φ(x)+λ(ρ-J(x))).
Then there exist a nonempty open set A⊆I and a positive number R with the following property: for every λ∈A and every C1 functional -Ψ:X→R with compact derivative, there exists δ>0 such that, for each μ∈[0,δ],
(15)Φ′(x)-λJ′(x)-μΨ′(x)=0
has at least three solutions in X whose norms are less than R.
Proposition 12 (see [34]).
Let X be a nonempty set and Φ,J two real functions on X. Assume that there are r>0 and x0,x1∈X such that
(16)Φ(x0)=J(x0)=0,Φ(x1)>r,supx∈Φ-1((-∞,r])J(x)<rJ(x1)Φ(x1).
Then, for each ρ, satisfying
(17)supx∈Φ-1((-∞,r])J(x)<ρ<rJ(x1)Φ(x1),
one has
(18)supλ≥0infx∈X(Φ(x)+λ(ρ-J(x)))<infx∈Xsupλ≥0(Φ(x)+λ(ρ-J(x))).
We always assume that the following conditions hold:
A(t)=(aij(t)) is a symmetric matrix with aij∈L∞([0,T]) for every t∈[0,T];
there exists a positive constant ν such that
(19)A(t)u·u≥ν|u|2,foreveryu∈RN,a.e.t∈[0,T];
for any i=1,2,…,N, and j=1,2,…,l, Iij(t) are nondecreasing in t∈R and
(20)Iij(t)t≥0,∀t∈R.
Define the Sobolev space by
(21)HT1≔{L2u:[0,T]⟶RN:uisabsolutelycontinuous,0000u(0)=u(T),andu˙∈L2(0,T)}
with the inner product
(22)〈u,v〉=∫0T(u(t),v(t))dt+∫0T(u˙(t),v˙(t))dt,hhhhhhhhhhhhhhhhhhhhhhhhhh∀u,v∈HT1,
where (·,·) denotes the inner product in RN. The corresponding norm is defined by
(23)∥u∥HT1=(∫0T|u(t)|2dt+∫0T|u˙(t)|2dt)1/2,∀u∈HT1.
For every u,v∈HT1, we define
(24)〈u,v〉1=∫0T(A(t)u(t),v(t))dt+∫0T(u˙(t),v˙(t))dt,∀u,v∈HT1
and observe that, using assumptions (A1) and (A2), (24) defines an inner product in HT1, whose corresponding norm is
(25)∥u∥≔〈u,u〉11/2.
A simple computation shows that
(26)(A(t)u,u)=∑i,j=1Nai,j(t)uiuj≤∑i,j=1N∥ai,j∥∞|u|2,
for every t∈[0,T] and u∈RN. So, putting together (A2) and (26), one has
(27)m∥u∥HT1≤∥u∥≤M∥u∥HT1,
where m=min{1,ν} and M=max{1,∑i,j=1N∥ai,j∥∞}; that is, the norm ∥·∥ is equivalent to (23). At this point, since (HT1,∥·∥HT1) is compactly embedded in C(0,T) (see [35]), due to (27) we claim that there exists a positive number k such that
(28)∥u∥C≤k∥u∥,
where k=(1/m)max{T,1/T} and ∥u∥C=maxt∈[0,T]|u(t)|.
If u∈HT1, then u is absolutely continuous and u˙∈L2(0,T;RN). In this case, Δu˙(t)=u˙(t+)-u˙(t-)=0 is not necessarily valid for every t∈(0,T) and the derivative u˙ may present some discontinuities. This leads to the impulsive effects.
Following the ideas of [15], take v∈HT1 and multiply the two sides of the equality
(29)-u¨+A(t)u-λ∇F(t,u)-μ∇G(t,u)=0
by v and integrate from 0 to T; we have
(30)∫0T[-u¨+A(t)u-λ∇F(t,u)-μ∇G(t,u)]vdt=0.
Moreover, combining u˙(0)-u˙(T)=0, one has
(31)∫0T(u¨(t),v(t))dt=∑j=0l∫tjtj+1(u¨(t),v(t))dt=∑j=0l((u˙(tj+1-),v(tj+1-))00000-(u˙(tj+),v(tj+)))-∑j=0l∫tjtj+1(u˙(t),v˙(t))dt=∑j=0l∑i=1N(u˙i(tj+1-)vi(tj+1-)00000000-u˙i(tj+)vi(tj+))-∑j=0l∫tjtj+1(u˙(t),v˙(t))dt=u˙(T)v(T)-u˙(0)v(0)-∑j=1l∑i=1NΔu˙i(tj)vi(tj)-∫0T(u˙(t),v˙(t))dt=-∑j=1l∑i=1NIij(ui(tj))vi(tj)-∫0T(u˙(t),v˙(t))dt.
Combining (30), we get
(32)∫0T(u˙(t),v˙(t))+(A(t)u(t),v(t))dt+∑j=1l∑i=1NIij(ui(tj))vi(tj)-λ∫0T∇F(t,u)vdt-μ∫0T∇G(t,u)vdt=0.
Considering the above equality, we introduce the following concept of a weak solution to problem (9).
Definition 13.
One says that a function u∈HT1 is a weak solution to problem (9) if
(33)∫0T(u˙(t),v˙(t))+(A(t)u(t),v(t))dt+∑j=1l∑i=1NIij(ui(tj))vi(tj)=λ∫0T∇F(t,u)vdt+μ∫0T∇G(t,u)vdt
holds for any v∈HT1.
In order to study problem (9), we will use the functionals Φ,J,Ψ:HT1→R defined by putting
(34)Φ(u)=12∥u∥2+∑j=1l∑i=1N∫0ui(tj)Iij(t)dt=ϕ(u)+ψ(u),J(u)≔∫0TF(t,u)dt,Ψ(u)≔∫0TG(t,u)dt,
respectively, for every u∈HT1; here,
(35)ϕ(u)=12∥u∥2,ψ(u)=∑j=1l∑i=1N∫0ui(tj)Iij(t)dt.
By the continuity of Iij, i=1,2,…,N, j=1,2,…,l, we get that functional Φ is a continuous Gâteaux differential functional whose Gâteaux derivative is the functional Φ′(u), given by
(36)〈Φ′(u),v〉=∫0T(A(t)u(t),v(t))+(u˙(t),v˙(t))dt+∑j=1l∑i=1NIij(ui(tj))vi(tj),
for any u,v∈HT1. On the other hand, condition (10) implies that J and Ψ are well defined, continuously Gâteaux differentiable in HT1. More precisely, their Gâteaux derivatives are
(37)〈J′(u),v〉=∫0T∇F(t,u)vdt,〈Ψ′(u),v〉=∫0T∇G(t,u)vdt,
respectively, for every u,v∈HT1. Put
(38)q=∑i,j=1N∥ai,j∥∞,λ1=inf{∥u∥2+2∑j=1l∑i=1N∫0ui(tj)Iij(t)dt2∫0TF(t,u(t))dt:u∈HT1,000000000∫0TF(t,u(t))dt>0∥u∥2+2∑j=1l∑i=1N∫0ui(tj)Iij(t)dt2∫0TF(t,u(t))dt},λ2=(max{0,limsup∥u∥→+∞2∫0TF(t,u(t))dt∥u∥2+2∑j=1l∑i=1N∫0ui(tj)Iij(t)dt,0000000000limsup|u|→02∫0TF(t,u(t))dt∥u∥2+2∑j=1l∑i=1N∫0ui(tj)Iij(t)dt})-1D=(T-tl)2t1tl2+t13tl2(tl2+tlT+T2)+(tl-t1)+T-tltl2+13tl2(T3-tl3)>0.
Then we have the following.
Theorem 14 (see [16, Theorem 3.1]).
Suppose that (A1)-(A2) and (I1) hold. Moreover, there exist a constant p>0 and constant vector ξ¯=(ξ¯1,…,ξ¯N)∈RN∖{0} such that
Then, for each compact interval [a,b]⊂(λ1,λ2), there exists R>0 with the following property: for every λ∈[a,b] and G∈C there exists δ>0 such that, for each μ∈[0,δ], problem (9) has at least three weak solutions whose norms are less than R.
Theorem 15 (see [16, Theorem 3.2]).
Suppose that (A1)-(A2) and (I1) hold and there exist h,q∈L1(0,T;R+), three positive constants c,d,α with c>kdDm/2, and α∈[0,2) such that
|F(t,u)|≤h(t)|u|α+q(t), for all u∈RN and a.e. t∈[0,T].
Then, there exist a nonempty open set A⊂[0,+∞) and a positive number R with the following property: for every λ∈A and every G∈C there exists δ>0 such that, for each μ∈[0,δ], problem (9) has at least three weak solutions whose norms are less than R.
As a special case of F(t,u), let F(t,u)=b(t)f(u), where b(t)∈L1(0,T;R)∖{0}, f(u)∈C1(RN;R). Then we get the following two corollaries.
Corollary 16 (see [16, Corollary 3.1]).
Suppose that (A1)-(A2) and (I1) hold. Moreover, there exist a constant p>0 and constant vector ξ¯=(ξ¯1,…,ξ¯N)∈RN∖{0} such that
Then, for each compact interval [a,b]⊂(λ1,λ2), there exists R>0 with the following property: for every λ∈[a,b] and G∈C there exists δ>0 such that, for each μ∈[0,δ], problem (9) has at least three weak solutions whose norms are less than R.
Corollary 17 (see [16, Corollary 3.2]).
Suppose that (A1)-(A2) and (I1) hold. Furthermore, there exist five positive constants c, d, α, β, and γ with c>kdDm/2 and α∈[0,2) such that
Then, there exist a nonempty open set A⊂[0,+∞) and a positive number R with the following property: for every λ∈A and every G∈C there exists δ>0 such that, for each μ∈[0,δ], problem (9) has at least three weak solutions whose norms are less than R.
Finally, we consider the existence of infinitely many solutions for a class of impulsive systems with periodic boundary conditions
(39)-u¨+A(t)u=∇W(t,u),a.e.t∈[0,T],Δ(u˙i(tj))=u˙i(tj+)-u˙i(tj-)=Iij(ui(tj)),hhhhhhhhhhhhi=1,2,…,N,j=1,2,…,l,u(0)-u(T)=u˙(0)-u˙(T)=0,
where A:[0,T]→RN×N is a continuous map from the interval [0,T] to the set of N-order symmetric matrices, T is a real positive number, u(t)=(u1(t),u2(t),…,uN(t)), tj, j=1,2,…,l, are the instants where the impulses occur and 0=t0<t1<t2<⋯<tl<tl+1=T, Iij:R→R(i=1,2,…,N,j=1,2,…,l) are continuous, and W:[0,T]×RN→R are measurable with respect to t, for every u∈RN, continuously differentiable in u, for almost every t∈[0,T], and satisfy the following standard summability condition:
(40)sup|u|≤bmax{|W(·,u)|,|∇W(·,u)|}∈L1(0,T),∀b>0.
Now we state our main results.
Theorem 18 (see [36, Theorem 1.1]).
Assume that W is even in u and the following conditions hold:
Iij(i=1,…,N,j=1,2,…,l) are odd and satisfy
(41)2∫0sIij(t)dt-Iij(s)s≥0,∫0sIij(t)dt≥0∀s∈R;
there exist constants aij,bij>0 and γij∈[0,1) such that
(42)|Iij(s)|≤aij+bij|s|γijfors∈R;
lim|u|→+∞(W(t,u)/u2)=+∞ uniformly for t∈[0,T];
W(t,0)≡0, 0≤W(t,u)=o(|u|2) as |u|→0 uniformly for t∈[0,T];
there exist constants α>1, 1<β<1+((α-1)/α), c1,c2>0, and L>0 such that for every t∈[0,T] and u∈RN with |u|≥L,
(43)(∇W(t,u),u)-2W(t,u)≥c1|u|α,|∇W(t,u)|≤c2|u|β.
Then problem (39) has infinitely many weak solutions.
Theorem 19 (see [36, Theorem 1.2]).
Assume that the following conditions hold:
Iij(i=1,2,…,N,j=1,2,…,l) are odd and satisfy Iij(s)s≥0 for any s∈R;
there exist constants bij>0 and γij∈[0,1) such that
(44)|Iij(s)|≤bij|s|γijfors∈R;
W(t,u)=(ρ/2)|u|2+F(t,u), where ρ is a positive constant and not a spectrum point of -(d2/dt2)+A, F is even in u, and F(t,0)≡0;
there exist δ1,δ2∈[1,2) with δ1<min1≤i≤N,1≤j≤l{γij}+1, and d1,d2>0 such that
(45)d1|u|δ1≤F(t,u),|∇F(t,u)|≤d2|u|δ2-1,hhhhhhhhhhhhhhh∀(t,u)∈[0,T]×RN.
Then problem (39) has infinitely many weak solutions.
Theorem 20 (see [36, Theorem 1.3]).
Assume that W is even in u, (S1)-(S2) and the following conditions hold:
W(t,0)≡0 for any t∈[0,T];
there are constants k1>0 and ζ1∈[1,2) with ζ1<min1≤i≤N,1≤j≤l{γij}+1 such that
(46)W(t,u)≥k1|u|ζ1,forany(t,u)∈[0,T]×RN;
there exist constants k2>0 and ζ2∈[1,2) such that
(47)|∇W(t,u)|≤k2|u|ζ2-1,forany(t,u)∈[0,T]×RN.
Then problem (39) has infinitely many weak solutions.
3. Periodic Solutions
Firstly, we consider periodic solutions for second-order regular impulsive problems with a perturbation:
(48)u′′(t)-Vu(t,u)=f(t),fort∈(sk-1,sk),Δu′(sk)=gk(u(sk)),
where k∈Z, u∈Rn, Δu′(sk)=u′(sk+)-u′(sk-) with u′(sk±)=limt→sk±, Vu(t,u)=graduV(t,u), f∈C(R,Rn), gk(u)=graduGk(u), Gk∈C1(Rn,Rn) for each k∈Z, and there exist m∈N and T∈R+ such that 0=s0<s1<⋯<sm=T, sk+m=sk+T, and gk+m≡gk for all k∈Z (i.e., gk is m-periodic in k).
Theorem 21 (see [20, Theorem 1]).
Assume that f is continuous and T-periodic and gk is continuous, m-periodic in k. Furthermore, assume that V and gk satisfy the following conditions:
V:R×Rn→R is continuous differentiable and T-periodic, and there exist positive constants b1,b2>0 such that
(49)b1|u|2≤-V(t,u)≤|u|2,∀(t,u)∈[0,T]×Rn;
-V(t,u)≤-Vu(t,u)u≤-2V(t,u) for all (t,u)∈[0,T]×Rn;
there exists μ>2 such that
(50)gk(u)u≤μGk(u)<0,foru∈Rn∖{0},k=1,2,…,m.
Then there exists b-f>0 such that, for any positive integer l, if bf,1≤b-fl-1/2, then problem (48) possesses at least one nonzero lT-periodic solution.
Theorem 22 (see [20, Theorem 1]).
Suppose that V is even about u and satisfies (V1)-(V2), f≡0, and gk is continuous, odd about u, m-periodic in k, and satisfies the following:
for all k=1,2,…,m,
(51)liminf|u|→∞Gk(u)|u|2≥0.
Then there exists Cg>0 such that if
(52)limsup|u|→0Gk(u)|u|2≤-Cg,
then problem (48) possesses at least lmn distinct pairs of lT-periodic solutions generated by impulses.
Theorem 23 (see [14, Theorem 1]).
Assume that f≡0 and gk is continuous, m-periodic in k satisfying (g1). Furthermore, assume that V satisfies the following condition:
V:R×Rn→R is continuous differentiable and T-periodic, and V(t,u)≥(1/2)Vu(t,u)u>0 for all t∈[0,T] and u∈Rn∖{0}.
Then problem (48) possesses one nonzero periodic solution generated by impulses.
Corollary 24 (see [14, Corollary 1]).
Assume that f≡0 and gk is continuous, m-periodic in k satisfying (g1). Furthermore, assume that V satisfies the following condition:
Vu(t,u)=αu+Wu(t,u) for all t∈[0,T] and u∈Rn, where α>2/μ for some μ>2 and W(t,u)≥(1/2)Wu(t,u)u>0 for all t∈[0,T] and u∈Rn∖{0}.
Then problem (48) possesses one nonzero periodic solution generated by impulses.
Theorem 25 (see [14, Theorem 2]).
Assume that f≡0 and V and gk satisfy the following conditions:
V(t,u)≥(1/2)Vu(t,u)u≥0 for all (t,u)∈[0,T]×Rn;
gk(u)=2u+wk(u), where wk=graduWk(u) and satisfies the fact that there exists μ>2 such that
(53)wk(u)u≤μWk(u)<0,foru∈Rn∖{0},hhhhhhhhhhhhhhhhhhhhhhk=1,2,…,m.
Then problem (48) possesses one nonzero periodic solution generated by impulses.
Now we discuss T-periodic solution to second-order impulsive problems with singularity as follows:
(54)u′′(t)-1uα(t)=e(t),a.e.t∈(0,T),Δu′(tj)=Ij(u(tj)),j=1,2,…,p-1,
where α≥1, e∈L1([0,T],R) is T-periodic, Δu′(tj)=u′(tj+)-u′(tj-) with u′(tj±)=limt→tj±u′(t); 0=t0<t1<t2<⋯<tp-1<tp=T and tj+p=tj+T; Ij∈C(R,R)(j=1,2,…,p-1) and Ij+p≡Ij.
Theorem 26 (see [18, Theorem 1.2]).
Assume that
e∈L1([0,T],R) is T-periodic and ∫0Te(t)dt≥0;
there exist two constants m,M such that, for any s∈R,
(55)m≤Ij(s)≤M,j=1,2,…,p-1,
where
(56)m≤M<-1p-1∫0Te(t)dt≤0.
Then problem (54) has at least a positive T-periodic solution.
Set
(57)HT1={u′L2u:R⟶R∣uisabsolutelycontinuous,000u′∈L2((0,T),R),andu(t)=u(t+T)fort∈R}
with the inner product
(58)(u,v)=∫0Tu(t)v(t)dt+∫0Tu′(t)v′(t)dt,∀u,v∈HT1.
The corresponding norm is defined by
(59)∥u∥HT1=(∫0T|u(t)|2dt+∫0T|u′(t)|2dt)1/2,∀u∈HT1.
Then HT1 is a Banach space (in fact it is a Hilbert space).
In order to study problem (54), for any λ∈(0,1), we consider the following modified problem:
(60)u′′(t)+fλ(u(t))=e(t),a.e.t∈(0,T),Δu′(tj)=Ij(u(tj)),j=1,2,…,p-1.
Following the ideas of [15], we introduce the following concept of a weak solution to problem (60).
Definition 27.
One says that a function u∈HT1 is a weak solution to problem (60) if
(61)∫0Tu′(t)v′(t)dt+∑j=1p-1Ij(u(tj))v(tj)-∫0Tfλ(u(t))v(t)dt+∫0Te(t)v(t)dt=0
holds for any v∈HT1.
Let Fλ∈C1(R,R) be defined by
(62)Fλ(s)=∫1sfλ(t)dt,
and consider the functional
(63)Φλ:HT1⟶R,
defined by
(64)Φλ(u)≔12∫0T|u′(t)|2dt+∑j=1p-1∫0u(tj)Ij(s)ds-∫0TFλ(u(t))dt+∫0Te(t)u(t)dt.
Clearly, Φλ is well defined on HT1, continuously Gâteaux differentiable functional whose Gâteaux derivative is the functional Φλ′(u), given by
(65)Φλ′(u)v=∫0Tu′(t)v′(t)dt+∑j=1p-1Ij(u(tj))v(tj)-∫0Tfλ(u(t))v(t)dt+∫0Te(t)v(t)dt,
for any v∈HT1. Moreover, it is easy to verify that Φλ is weakly lower semicontinuous. Indeed, if {un}⊂HT1, u∈HT1, and un⇀u, then {un} converges uniformly to u on [0,T] and un→u on L2([0,T]), and combining the fact that liminfn→∞∥un∥HT1≥∥un∥HT1, we have
(66)liminfn→∞Φλ(un)=liminfn→∞(12∥un∥HT12-12∫0T|un(t)|2dt000000000+∑j=1p-1∫0un(tj)Ij(s)ds000000000-∫0TFλ(un(t))dt000000000+∫0Te(t)un(t)dt)≥12∫0T|u′(t)|2dt+∑j=1p-1∫0u(tj)Ij(s)ds-∫0TFλ(u(t))dt+∫0Te(t)u(t)dt=Φλ(u).
By the standard discussion, the critical points of Φλ are the weak solutions of problem (1); see [15, 16].
The following version of the mountain pass theorem will be used in our argument.
Theorem 28 (see Theorem 4.10, [35]).
Let X be a Banach space and let φ∈C1(X,R). Assume that there exist x0,x1∈X and a bounded open neighborhood Ω of x0 such that x1∈X∖Ω¯ and
(67)max{φ(x0),φ(x1)}<infx∈∂Ωφ(x).
Let
(68)Γ={h∈C([0,1],X):h(0)=x0,h(1)=x1},c=infh∈Γmaxs∈[0,1]φ(h(s)).
If φ satisfies the (PS) condition, that is, a sequence {un} in X satisfying φ(un) is bounded and φ′(un)→0 as n→∞ has a convergent subsequence, then c is a critical value of φ and c>max{φ(x0),φ(x1)}.
If Ij(u(tj))≡bj (j=1,2,…,p-1), then we have the following result of problem (54).
Theorem 29 (see [17, Theorem 1.2]).
Assume that e∈L1([0,T],R) is T-periodic. Then problem (54) has a positive T-periodic weak solution u∈HT1 if and only if ∑j=1p-1bj+∫0Te(t)dt<0.
Remark 30.
From Theorem 29 we can see that if ∫0Te(t)dt≥0 but ∑j=1p-1bj+∫0Te(t)dt<0, then problem (54) still admits a positive T-periodic solution. This shows that the existence of positive T-periodic solution to problem (54) depends on the forced term e and impulsive functions bj together, not single one.
4. Homoclinic Solutions
In this section, we are interested in the existence of homoclinic solutions for problem (48).
First we consider the case of f≡0. When Vu(t,u) is asymptotically linear, we show that there exists a nonzero homoclinic solution as the limit of a sequence 2lT-periodic solutions as l goes to ∞, and when Vu(t,u) is sublinear, we attain a nonzero homoclinic solution in the same way by strengthening the effect of impulses. The results are as follows.
Theorem 31 (see [14, Theorem 3]).
Supposing that f≡0, V is T-periodic in t and satisfies condition (V4), and gk satisfies condition (g1), then problem (48) possesses at least one nonzero homoclinic solution generated by impulses.
Theorem 32 (see [14, Theorem 4]).
Supposing that f≡0, V is T-periodic in t and satisfies condition (V5), and gk satisfies condition (g3), then problem (48) possesses at least one nonzero homoclinic solution generated by impulses.
If n=1 and gk(u)≡I(u) for all k=1,2,…,m, then we have the following result for problem (48) when Vu satisfies the superlinear condition at infinity on u.
Theorem 33 (see [13, Theorem 1.1]).
Assume that f≡0 and the following conditions hold:
there exists a positive number T such that
(69)Vu(t+T,u)=Vu(t,u),V(t+T,u)=V(t,u)hhhhhhhhhhhhhhhhhhhhhhhhhllhh∀(t,u)∈R2;
limu→0Vu(t,u)u=0 uniformly for t∈R;
there exists a μ>2 such that
(70)uVu(t,u)≥μV(t,u)>0,∀(t,u)∈R×R∖{0};
there exist constants a0>0 and a1>0 such that
(71)V(t,u)≥a0|u|μ,∀|u|≥1,t∈R;V(t,u)≤a0|u|μ,∀|u|≤1,t∈R;
there exists a constant b, with 0<b<((μ-2)/(μ+2)Tm), such that
(72)|I(u)|≤b|u|,2∫0uI(t)dt-I(u)u≤0.
Then there exists a nontrivial weak homoclinic solution to problem (48).
The proof of Theorem 33 is based on the following version of Mountain pass theorem.
Theorem 34 (see [37]).
Let E be a Banach space and let φ∈C1(E,R), e∈E, r>0 be such that ∥e∥>r and b:=inf∥y∥=rφ(y)>φ(0)≥φ(e). Let
(73)Γ={h∈C([0,1],X):h(0)=x0,h(1)=x1},c≔infh∈Γmaxs∈[0,1]φ(h(s)).
Then, for each ϵ>0, δ>0, there exists y∈E such that
c-2ϵ≤φ(y)≤c+2ϵ;
dist(y,E)≤2δ;
∥φ′(y)∥≤8ϵ/δ.
Next we consider the case of f≢0 for problem (48).
Theorem 35 (see [20, Theorem 3]).
Assume that V satisfies conditions (V1)-(V2) and gk satisfies (g1). Furthermore, gk satisfies the following condition:
gk∈C(R) is m-periodic in k, and lim|u|→0(gk(u)/|u|)=0 for k=1,2,…,m.
Then there exists b-f>0 such that if ∫R|f(t)|2dt≤b-f2, then problem (48) possesses at least one nonzero homoclinic solution.
If V satisfies some other conditions, then we still obtain the existence result for problem (48).
Theorem 36 (see [19, Theorem 1]).
Assume that the following conditions hold:
V(t,u)=-K(t,u)+W(t,u), where K, W∈C1(R×Rn,R), and is T-periodic in its first variable;
there exist constants a1>0 and γ∈(1,2] such that for all (t,u)∈[0,T]×Rn(74)K(t,0)≡0,K(t,u)≥a1|u|γ;
there exists θ∈(1,2] such that
(75)K(t,u)≤Ku(t,u)u≤θK(t,u),∀(t,u)∈[0,T]×Rn;
W(t,0)=0, Wu(t,u)=o(|u|), as |u|→0 uniformly for t∈[0,T];
there exists r>2 and b1>0 such that
(76)W(t,u)≤b1|u|r,∀(t,u)∈[0,T]×Rn;
there exist μ>1, a2>0, and β(t)∈L1(R,R+) such that μ>r-γ and
(77)Wu(t,u)u-2W(t,u)≥a2|u|μ-β(t),hhhhhhhhhhhhhl∀(t,u)∈[0,T]×Rn;
W(t,u)/|u|2→+∞ as |u|→+∞ uniformly in t∈[0,T];
Gk(0)=0 and lim|u|→0gk(u)|u|=0 for k=1,2,…,m;
there exists b2>0 such that Gk(u)≤b2|u|r, u∈Rn, k=1,2,…,m;
Then problem (48) possesses at least one nonzero homoclinic solution.
Conflict of Interests
The authors declare that there is no conflict of interests regarding the publication of this paper.
Acknowledgments
J. Sun was supported by the National Natural Science Foundation of China (Grant nos. 11201270 and 11271372), Shandong Natural Science Foundation (Grant no. ZR2012AQ010), and Young Teacher Support Program of Shandong University of Technology.
CarterT. E.Optimal impulsive space trajectories based on linear equations199170227729710.1007/BF00940627MR1120955ZBL0732.49025CarterT. E.Necessary and sufficient conditions for optimal impulsive rendezvous with linear equations of motion200010321922710.1023/A:1008376427023MR1809484ZBL0980.93058ChuJ.NietoJ. J.Impulsive periodic solutions of first-order singular differential equations200840114315010.1112/blms/bdm110MR2409187ZBL1144.34016LiW.HuoH.Global attractivity of positive periodic solutions for an impulsive delay periodic model of respiratory dynamics2005174222723810.1016/j.cam.2004.04.010MR2106438ZBL1070.34089LiC.LiaoX.YangX.HuangT.Impulsive stabilization and synchronization of a class of chaotic delay systems200515904310310.1063/1.2102107MR2194910ZBL1144.37371PradoA. F. B. A.Bi-impulsive control to build a satellite constellation200552169175MR2140558ZBL1128.70015AgarwalR. P.FrancoD.O'ReganD.Singular boundary value problems for first and second order impulsive differential equations2005691-2839610.1007/s00010-004-2735-9MR2126188ZBL1073.34025AhmadB.NietoJ. J.Existence and approximation of solutions for a class of nonlinear impulsive functional differential equations with anti-periodic boundary conditions200869103291329810.1016/j.na.2007.09.018MR2450538ZBL1158.34049LakshmikanthamV.BainovD. D.SimeonovP. S.19896SingaporeWorld Scientific10.1142/0906MR1082551LiJ.NietoJ. J.ShenJ.Impulsive periodic boundary value problems of first-order differential equations2007325122623610.1016/j.jmaa.2005.04.005MR2273040ZBL1110.34019NietoJ. J.Basic theory for nonresonance impulsive periodic problems of first order199712525992604MR1428357ZBL0870.34009SamoilenkoA. M.PerestyukN. A.199514SingaporeWorld Scientific10.1142/9789812798664MR1355787FangH.DuanH.Existence of nontrivial weak homoclinic orbits for second-order impulsive differential equations20122012, article 1381310.1186/1687-2770-2012-138MR3037631ZBL1283.34024HanX.ZhangH.Periodic and homoclinic solutions generated by impulses for asymptotically linear and sublinear Hamiltonian system201123551531154110.1016/j.cam.2010.08.040MR2728109ZBL1211.34008NietoJ. J.O'ReganD.Variational approach to impulsive differential equations200910268069010.1016/j.nonrwa.2007.10.022MR2474254ZBL1167.34318SunJ.ChenH.NietoJ. J.Otero-NovoaM.The multiplicity of solutions for perturbed second-order Hamiltonian systems with impulsive effects201072124575458610.1016/j.na.2010.02.034MR2639205ZBL1198.34036SunJ.ChuJ.A suffcient and necessary condition on T-periodic solution for a singular problem with impulses201494110SunJ.ChuJ.ChenH.Periodic solution generated by impulses for singular differential equations2013404256256910.1016/j.jmaa.2013.03.036MR3045195XieJ.LuoZ.ChenG.Homoclinic solutions for a class of the second-order impulsive Hamiltonian systems20132013810.1155/2013/583107583107MR3093764ZhangH.LiZ.Periodic and homoclinic solutions generated by impulses2011121395110.1016/j.nonrwa.2010.05.034MR2728662ZBL1225.34019ZhangZ.YuanR.An application of variational methods to Dirichlet boundary value problem with impulses201011115516210.1016/j.nonrwa.2008.10.044MR2570535ZBL1191.34039ZhouJ.LiY.Existence and multiplicity of solutions for some Dirichlet problems with impulsive effects2009717-82856286510.1016/j.na.2009.01.140MR2532812ZBL1175.34035AgarwalR. P.BhaskarT. G.PereraK.Some results for impulsive problems via Morse theory2014409275275910.1016/j.jmaa.2013.07.036MR3103194ChenH.SunJ.An application of variational method to second-order impulsive differential equation on the half-line201021751863186910.1016/j.amc.2010.06.040MR2727931ZBL1213.34020SunJ.ChenH.Variational method to the impulsive equation with Neumann boundary conditions200920091710.1155/2009/316812316812MR2559028ZBL1184.34039SunJ.ChenH.Multiplicity of solutions for a class of impulsive differential equations with Dirichlet boundary conditions via variant fountain theorems20101154062407110.1016/j.nonrwa.2010.03.012MR2683856ZBL1208.34031TianY.GeW.Applications of variational methods to boundary-value problem for impulsive differential equations200851250952710.1017/S0013091506001532MR2465922ZBL1163.34015TianY.GeW.Variational methods to Sturm-Liouville boundary value problem for impulsive differential equations201072127728710.1016/j.na.2009.06.051MR2574937ZBL1191.34038TianY.GeW.Multiple solutions of impulsive Sturm-Liouville boundary value problem via lower and upper solutions and variational methods2012387247548910.1016/j.jmaa.2011.08.042MR2853121ZBL1241.34032ZhangD.DaiB.Existence of solutions for nonlinear impulsive differential equations with Dirichlet boundary conditions2011535-61154116110.1016/j.mcm.2010.11.082MR2769486ZBL1217.34043ZhangH.LiZ.Variational approach to impulsive differential equations with periodic boundary conditions2010111677810.1016/j.nonrwa.2008.10.016MR2570525ZBL1186.34089RicceriB.A further three critical points theorem20097194151415710.1016/j.na.2009.02.074MR2536320ZBL1187.47057RicceriB.A three critical points theorem revisited20097093084308910.1016/j.na.2008.04.010MR2503052ZBL1214.47079RicceriB.Existence of three solutions for a class of elliptic eigenvalue problems20003211–131485149410.1016/S0895-7177(00)00220-XMR1800671ZBL0970.35089MawhinJ.WillemM.198974New York, NY, USASpringer10.1007/978-1-4757-2061-7MR982267SunJ.ChenH.NietoJ. J.Infinitely many solutions for second-order Hamiltonian system with impulsive effects2011541-254455510.1016/j.mcm.2011.02.044MR2801909ZBL1225.37070BrezisH.NirenbergL.Remarks on finding critical points1991448-993996310.1002/cpa.3160440808MR1127041ZBL0751.58006