
We consider an autonomous ordinary differential equation that admits a heteroclinic loop. The unperturbed heteroclinic loop consists of two degenerate heteroclinic orbits γ1 and γ2. We assume the variational equation along the degenerate heteroclinic orbit γi has di(di>1,i=1,2) linearly independent bounded solutions. Moreover, the splitting indices of the unperturbed heteroclinic orbits are s and −s (s≥0), respectively. In this paper, we study the persistence of the heteroclinic loop under periodic perturbation. Using the method of Lyapunov-Schmidt reduction and exponential dichotomies, we obtained the bifurcation function, which is defined from Rd1+d2+2 to Rd1+d2. Under some conditions, the perturbed system can have a heteroclinic loop near the unperturbed heteroclinic loop.
Citation: Bin Long, Shanshan Xu. Persistence of the heteroclinic loop under periodic perturbation[J]. Electronic Research Archive, 2023, 31(2): 1089-1105. doi: 10.3934/era.2023054
[1] | Haijun Wang, Jun Pan, Guiyao Ke . Multitudinous potential homoclinic and heteroclinic orbits seized. Electronic Research Archive, 2024, 32(2): 1003-1016. doi: 10.3934/era.2024049 |
[2] | Jun Pan, Haijun Wang, Feiyu Hu . Revealing asymmetric homoclinic and heteroclinic orbits. Electronic Research Archive, 2025, 33(3): 1337-1350. doi: 10.3934/era.2025061 |
[3] | Guowei Yu . Heteroclinic orbits between static classes of time periodic Tonelli Lagrangian systems. Electronic Research Archive, 2022, 30(6): 2283-2302. doi: 10.3934/era.2022116 |
[4] | Minzhi Wei . Existence of traveling waves in a delayed convecting shallow water fluid model. Electronic Research Archive, 2023, 31(11): 6803-6819. doi: 10.3934/era.2023343 |
[5] | Anna Gołȩbiewska, Marta Kowalczyk, Sławomir Rybicki, Piotr Stefaniak . Periodic solutions to symmetric Newtonian systems in neighborhoods of orbits of equilibria. Electronic Research Archive, 2022, 30(5): 1691-1707. doi: 10.3934/era.2022085 |
[6] | Yang Song, Beiyan Yang, Jimin Wang . Stability analysis and security control of nonlinear singular semi-Markov jump systems. Electronic Research Archive, 2025, 33(1): 1-25. doi: 10.3934/era.2025001 |
[7] | Xianjun Wang, Huaguang Gu, Bo Lu . Big homoclinic orbit bifurcation underlying post-inhibitory rebound spike and a novel threshold curve of a neuron. Electronic Research Archive, 2021, 29(5): 2987-3015. doi: 10.3934/era.2021023 |
[8] | Feng Wang, Taotao Li . Dynamics of a stochastic epidemic model with information intervention and vertical transmission. Electronic Research Archive, 2024, 32(6): 3700-3727. doi: 10.3934/era.2024168 |
[9] | Xiaochen Mao, Weijie Ding, Xiangyu Zhou, Song Wang, Xingyong Li . Complexity in time-delay networks of multiple interacting neural groups. Electronic Research Archive, 2021, 29(5): 2973-2985. doi: 10.3934/era.2021022 |
[10] | Yan Xie, Rui Zhu, Xiaolong Tan, Yuan Chai . Inhibition of absence seizures in a reduced corticothalamic circuit via closed-loop control. Electronic Research Archive, 2023, 31(5): 2651-2666. doi: 10.3934/era.2023134 |
We consider an autonomous ordinary differential equation that admits a heteroclinic loop. The unperturbed heteroclinic loop consists of two degenerate heteroclinic orbits γ1 and γ2. We assume the variational equation along the degenerate heteroclinic orbit γi has di(di>1,i=1,2) linearly independent bounded solutions. Moreover, the splitting indices of the unperturbed heteroclinic orbits are s and −s (s≥0), respectively. In this paper, we study the persistence of the heteroclinic loop under periodic perturbation. Using the method of Lyapunov-Schmidt reduction and exponential dichotomies, we obtained the bifurcation function, which is defined from Rd1+d2+2 to Rd1+d2. Under some conditions, the perturbed system can have a heteroclinic loop near the unperturbed heteroclinic loop.
The problems in homoclinic or heteroclinic bifurcation are critical in dynamic systems because they may have some complex dynamic behavior, such as chaotic motions [1]. Homoclinic and heteroclinic orbits are important invariant sets. The homoclinic orbit tends asymptotically to the same hyperbolic equilibrium along stable and unstable manifolds. However, the heteroclinic orbit tends asymptotically to two different hyperbolic equilibria along the stable and unstable manifolds. A heteroclinic loop consists of two saddles connecting two heteroclinic orbits. A numerical simulation reveals that the Lorenz equation has a heteroclinic loop when σ=10, r≈40.375 and b≈2.623 [2]. The heteroclinic loop is equidimensional if the two saddles have the same dimension of the unstable manifold. Otherwise, it is heterodimensional loop [3]. This elementary phenomenon occurs in any dimension larger than two, and is one of the primary mechanisms for non-hyperbolicity. In addition, the existence of the heteroclinic loop is often related to the traveling wave solutions of the reaction-diffusion equation.
In [4], Han et al. considered quadratic Hamiltonian systems with a heteroclinic loop under polynomial perturbations. Using the Melnikov function, the authors found three limit cycles near the heteroclinic loop. Later, Sun, Han, and Yang extended the theory for a heteroclinic loop with a cusp in [5]. Chen, Oksasoglu, and Wang considered a heteroclinic loop under periodic perturbation on the plane [6]. They proved three types of dynamic behavior near the heteroclinic loop under periodic perturbation. One of which with strange attractors admitting SRB measures representing chaos. More complicated dynamic behavior, such as strange attractors and horseshoes near the heteroclinic loop with periodic perturbation see, [7] and [8].
Chow, Deng, and Terman [9] investigated the homoclinic or periodic orbit bifurcated from a heteroclinic loop based on the method developed by Shilnikov. In 1998, Zhu and Xia [10] established a coordinate system in a neighborhood of a heteroclinic loop. They studied the bifurcation of the heteroclinic loop using the coordinate systems near the heteroclinic loop. Moreover, Rademacher [11] studied the homoclinic orbit bifurcated from a codimension 1 and 2 heteroclinic loops by Lin's method [12]. In [13], Geng, Wang, and Liu investigated the bifurcation of a heterodimensional loop using the local coordinate system. They assumed the unperturbed equation has a heteroclinic loop in R4 that the splitting indices of the unperturbed heteroclinic orbits are 1 and −1. They obtained the persistence condition for the heterodimensional loop. For more research results regarding the bifurcation of the heteroclinic loop see [14].
We let d,d≥1, denote the number of the bounded solutions of the variational equation along the heteroclinic orbit. If d=1, the homoclinic or heteroclinic orbit is nondegenerate; otherwise, it is degenerate [15], which means, along the orbit, the intersection of the spaces tangent to the stable and unstable manifolds of the equilibrium has a d dimensional subspace. Hence, parameter d describes the degeneration of the homoclinic or heteroclinic orbit.
The primary purpose of this paper is to extend the theory of [13,14] for heteroclinic loop bifurcation. We consider an autonomous ordinary differential equation that admits a heteroclinic loop in Rn. The unperturbed heteroclinic loop consists of two degenerate heteroclinic orbits. Furthermore, the splitting index of the unperturbed heteroclinic orbits can be arbitrary. We investigate the bifurcation of the heterodimensional loop under periodic perturbation using the Lyapunov-Schmidt reduction method. We start with the following equation:
˙x(t)=f(x(t)), | (1.1) |
and its periodic perturbed equation is as follows:
˙x(t)=f(x(t))+Σ2j=1μjgj(x(t),μ,t), | (1.2) |
where x∈Rn, μ=(μ1,μ2)∈R2, and we make the following assumptions:
(H1) f∈C3.
(H2) p+ and p− are the two distinct hyperbolic equilibria of Eq (1.1). Namely, f(p±)=0 and the eigenvalues of Df(p±) lie off the imaginary axis, where D denotes the derivative operator.
(H3) Equation (1.1) has two heteroclinic solutions γ1(t) and γ2(t), which are asymptotic to the equilibrium p+ and p−, respectively. That is, ˙γi(t)=f(γi(t)), i=1,2, and
limt→+∞γ1(t)=p+,limt→−∞γ1(t)=p−, |
limt→+∞γ2(t)=p−,limt→−∞γ2(t)=p+. |
(H4) gj∈C3, gj(p±,μ,t)=0, gj(x,0,t)=0 and gj(x,μ,t+2)=gj(x,μ,t).
(H5) dim(Ws(p+))=d+ and dim(Ws(p−))=d−, where Ws(p+) and Ws(p−) are the stable manifold of the equilibrium p+ and p−, respectively.
(H6)
dim(Tγ1(0)Ws(p+)⋂Tγ1(0)Wu(p−))=d1 |
and
dim(Tγ2(0)Ws(p−)⋂Tγ2(0)Wu(p+))=d2, |
where Tγi(0)Ws/u(p±) is the tangent spaces of the corresponding invariant manifolds at γi(0) and di>1, i=1,2.
By (H3) and (H6), we know unperturbed Eq (1.1) has a heteroclinic loop Γ (see Figure 1), where
Γ={p−}∪{γ1(t):t∈R}∪{p+}∪{γ2(t):t∈R} |
.
By (H5), we know that d+ and d− can be arbitrary. Thus, the unperturbed Eq (1.1) has a heterodimensional loop. We provide conditions for the persistence of the heterodimensional loop under periodic perturbation. The structure of the paper is as follows. We present some background on the Lyapunov-Schmidt reduction and Lin's method in Section 2. Section 3 details the notations for the fundamental matrix of the variational equation along the heteroclinic orbit γi(t) and the main result. Section 4 provides proof of the main result. The bifurcation function is obtained using the functional analytic method. We construct some solutions near the unperturbed heteroclinic loop, which can have a gap at t=0, and glue those solutions at t=0. Thus, the bifurcation function can be obtained. Hence, under some conditions, some solutions near the unperturbed heteroclinic loop can constitute a heteroclinic loop for a perturbed system.
Many problems in bifurcation theory can be changed by solving the zeros of an operator equation in some Banach space. Sometimes, the corresponding operator is not invertible, making it difficult to solve. However, this problem can equivalently transform the operator equation into an equation in a low-dimensional space using the Lyapunov-Schmidt reduction method (see [16]). Therefore, this method is very effective, especially in studying homoclinic or heteroclinic bifurcation.
Lin's method [17] is an implementation of the Lyapunov-Schmidt reduction method to construct solutions near the unperturbed heteroclinic orbit. The idea of Lin's method originated from the work by Chow, Hale, and Mallet-Paret [18] using the function space approach to construct piecewise continuous solutions approximating the unperturbed homoclinic orbit. The bifurcation function can be obtained using these solutions, and the zeros of the bifurcation function correspond to solutions in the homoclinic or heteroclinic bifurcation problems. Later, Palmer [19], Hale and Lin [20] extended the methods to Rn and the functional differential equation. Lin used the function space approach to construct solutions near the heteroclinic chain [12]. He assumed that heteroclinic orbits in the chain all have the same index. In the 1990s, Gruendler [21,22] generalized the method to the case of degenerate homoclinic bifurcation problems.
Next, we introduce an application of the Lyapunov-Schmidt reduction method, known as the Fredholm alternative property for linear differential equations. We consider the following nonhomogeneous linear differential equation:
˙y(t)=A(t)y(t)+h(t), | (2.1) |
where y∈Rn, A(t) vary continuously with t∈R and h(t) is bounded and continuous on t∈R. We assume that the homogeneous differential equation ˙y(t)=A(t)y(t) has exponential dichotomies on R+ and R−, respectively. Then, M>0, K0>0, and projections P and Q exist, such that
|U(t)PU−1(s)|≤K0e2M(s−t), 0⩽s⩽t,|U(t)(I−P)U−1(s)|≤K0e2M(t−s), 0⩽t⩽s,|U(t)(I−Q)U−1(s)|≤K0e2M(t−s), t⩽s⩽0,|U(t)QU−1(s)|≤K0e2M(s−t), s⩽t⩽0, | (2.2) |
where U(t) is the fundamental matrix. We define the Banach spaces as follows:
Zr={z∈Cr(R,Rn):max0≤j≤r{supt∈R|Djz(t)|eM|t|}<∞}, |
with the norm ||z||r=max0≤j≤r{supt∈R|Djz(t)|eM|t|}, |D0z(t)| indicates |z(t)|. We let the linear operator L:Z1→Z0 be defined by
L(y):=˙y−A(t)y. | (2.3) |
The adjoint operator for L is
L∗(ψ):=˙ψ+(A(t))Tψ, | (2.4) |
where (A(t))T denotes the transpose of matrix A(t)). By the definition of the linear operator L and the exponential dichotomy, we know that
dimKer(L)=dim(Ran(P)∩Ran(I−Q)),dimKer(L∗)=dim(Ran(I−PT)∩Ran(QT)). |
If dimKer(L∗)=d and ψ1(t),...,ψd(t) are the orthonormal unit bases of Ker(L∗), we define a projection operator Π:Z0→Z0 as follows
Π(h)(t)=d∑i=1ψi(t)∫∞−∞⟨ψTi(t),h(t)⟩dt. | (2.5) |
By the method of the Lyapunov-Schmidt reduction, Eq (2.1) is equivalent to the following system
˙y=A(t)y+(I−Π)h(t), | (2.6) |
Πh(t)=0. | (2.7) |
By the definition of Π, Ran(I−Π)=RanL. We can first solve Eq (2.6) for y∈Z1, and the bifurcation equations are obtained by Eq (2.7). That is,
d∑i=1ψi(t)∫∞−∞⟨ψTi(t),h(t)⟩dt=0,for allψi∈Ker(L∗). | (2.8) |
Thus, Eq (2.1) has a bounded solution y(t) if and only if Eq (2.8) holds.
The variational equation of (1.1) along the heteroclinic orbit γi is:
˙u(t)=Df(γi(t))u(t). | (3.1) |
From (H6), we know that Eq (3.1) has di(di>1) linearly independent bounded solutions, i=1,2. Based on Sacker's definition [23], we can define the splitting index S(γi) for the unperturbed heteroclinic orbit γi, as follows:
S(γ1)=d+−d−=s,S(γ2)=d−−d+=−s. | (3.2) |
By (H3) and the exponential dichotomy roughness theorem, we know that the variational Eq (3.1) has two-side exponential dichotomies. We let Ui be the fundamental matrix of Eq (3.1). Then, M>0, K0>0, projections Pi and Qi exist, such that
|Ui(t)PiU−1i(s)|≤K0e2M(s−t), 0⩽s⩽t,|Ui(t)(I−Pi)U−1i(s)|≤K0e2M(t−s), 0⩽t⩽s,|Ui(t)(I−Qi)U−1i(s)|≤K0e2M(t−s), t⩽s⩽0,|Ui(t)QiU−1i(s)|≤K0e2M(s−t), s⩽t⩽0, | (3.3) |
where I is the n×n unit matrix. We let the linear operator Li:Z1→Z0 be defined by
Li(u):=˙u−Df(γi(t))u. | (3.4) |
Further, the adjoint operator for Li is
L∗i(ψ):=˙ψ+(Df(γi(t)))Tψ. | (3.5) |
We let U−1i denote the inverse of Ui. Then we have U−1iUi=I. Differentiating U−1i(t)Ui(t)=I with respect to t, we obtain
U−1i˙Ui+˙Ui−1Ui=0 |
hence,
˙Ui−1=−U−1i˙UiU−1i=−U−1iDf(γi). |
Therefore, we have
(˙Ui−1)T=−Df(γi)T(U−1i)T. |
We know that (U−1i)T is a matrix solution of the adjoint equation of (3.1). Taking the transpose in Eq (2.2), it is apparent that the adjoint equation of (3.1) also has exponential dichotomy on R+ with projection I−PTi, and on R− with projection I−QTi, respectively.
By the definition of the linear operator Li and the exponential dichotomy, we know that
dimKer(L1)=dim(Ran(P1)∩Ran(I−Q1))=dim(Tγ1(0)Ws(p+)∩Tγ1(0)Wu(p−))=d1,dimKer(L2)=dim(Ran(P2)∩Ran(I−Q2))=dim(Tγ2(0)Ws(p−)∩Tγ2(0)Wu(p+))=d2,dimKer(L∗i)=dim(Ran(I−PTi)∩Ran(QTi)). |
From the theory of homoclinic bifurcation, the linear operators L1 and L2 are Fredholm operators, and the index of the Fredholm operator Li is
indexLi=dimKer(Li)−codimRan(Li). |
If dimKer(L∗i)=d∗i,i=1,2, then we have
indexL1=d1−d∗1=d+−d−=S(γ1)=s,indexL2=d2−d∗2=d–d+=S(γ2)=−s. |
In addition, if ui1(t),...,uidi−1(t),˙γi(t) are the orthonormal unit bases of Ker(Li), φ1(t),...,φd1−s(t) are the orthonormal unit bases of Ker(L∗1) and ψ1(t),...,ψd2+s(t) are the orthonormal unit bases of Ker(L∗2), then define
a1i,k(α1)=∫+∞−∞⟨ψTi(s),gk(γ1(s),μ,s+α1)⟩ds,b1i,pq=∫+∞−∞⟨ψTi(s),D11f(γ1(s))u1p(s)u1q(s)⟩ds, |
where i=1,...,d1−s, p,q=1,...,d1−1, and k=1,2. Moreover,
a2j,k(α2)=∫+∞−∞⟨φTi(s),gk(γ2(s),μ,s+α2)⟩dt,b2j,mn=∫+∞−∞⟨φTi(s),D11f(γ2(s))u2m(s)u2n(s)⟩ds, |
where j=1,...,d2+s, m,n=1,...,d2−1, and k=1,2. Using those notations, we let M1:Rd1−1×R2×R→Rd1−s be given by
M1(β1,μ,α1)=(M11(β1,μ,α1),...,M1d1−s(β1,μ,α1)), |
and
M1i(β1,μ,α1)=2∑k=1a1i,k(α1)μk+12d1−1∑p=1d1−1∑q=1b1i,pqβ1pβ1q, |
where i=1,...,d1−s,β1=(β11,...,β1d1−1).
We let M2:Rd2−1×R2×R→Rd2+s be given by
M2(β2,μ,α2)=(M21(β2,μ,α2),...,M2d2+s(β2,μ,α2)), |
and
M2j(β2,μ,α2)=2∑k=1a2j,k(α2)μk+12d2−1∑m=1d2−1∑n=1b2j,mnβ2mβ2n, |
where j=1,...,d2+s and β2=(β21,...,β2d2−1). Further, we let M:Rd1+d2−2×R2×R2→Rd1−s×Rd2+s be given by
M(β,μ,α)=(M1(β1,μ,α1),M2(β2,μ,α2)), | (3.6) |
where β=(β1,β2),α=(α1,α2).
We can state the main result as follows:
Theorem 1. Assume that (H1)−(H5) hold. Let M(β,μ,α) be as in Eq (3.6). If there are some points (β0,μ0,α0)∈Rd1+d2−2×R2×R2, such that
M(β0,μ0,α0)=0 |
and
D(β,μ)M(β0,μ0,α0) |
is a nonsingular (d1+d2)×(d1+d2) matrix, then there exists an open interval I containing origin, the C1 function κ2:I→R2, and the heteroclinic solutions x1(ε,t), x2(ε,t) of the Eq (1.2) with μ=ε2(μ0+κ2(ε)), where ε∈I∖{0}, x1(ε,t) and x2(ε,t) are located near the heteroclinic orbits γ1 and γ2, such that x1(ε,t), x2(ε,t), p+ and p− can constitute a heteroclinic loop Γε.
The proof of Theorem 1 is performed in Section 4. The heteroclinic loop Γε as illustrated in Figure 2.
By (H2), we know the unperturbed Eq (1.1) has a heteroclinic loop Γ. In this section, we find conditions such that the perturbed Eq (1.2) have a heteroclinic loop Γμ with sufficiently small μ. For i=1 or i=2, we suppose xi(t) is a solution of Eq (1.2). With the change of variable
xi(t+αi)=γi(t)+zi(t), | (4.1) |
Equation (1.2) can be transformed into
˙zi=Df(γi)zi+˜g(zi,μ,αi), | (4.2) |
where
˜g(zi,μ,αi)(t)=f(γi(t)+zi(t))−f(γi(t))−Df(γi(t))zi(t)+Σ2j=1μjgj(γi(t)+zi(t),μ,t+αi). | (4.3) |
By direct calculation, we have
(i)˜g(0,0,αi)=0;D1˜g(0,0,αi)=0;(ii)D11˜g(0,0,αi)=D11f(γi);(iii)∂˜g∂μj(0,0,αi)(t)=gj(γi,0,t+αi), |
where Di and Dij denote the derivative of the multivariate function concerning its i-th and i and j-th variables, respectively.
Because we only consider the Eq (1.1) under a small periodic perturbed equation, we suppose μ∈¯B1(0,δ)⊆R2, where ¯B1(0,δ) is a closed set with radius δ>0 centered at the origin. Moreover, we have the following property regarding the function ˜g.
Lemma 1. The function ˜g(⋅,μ,αi):Z1ׯB1(0,δ)×R↦Z0.
Proof. For i=1 or i=2, we let zi∈Z1 be given. We can choose a closed set B such that zi(t),γi(t),zi(t)+γi(t) and p±+zi(t)+γi(t) are all ∈B for t∈R. According to smoothness of f,gj∈C3 and gj is periodic about t. We can choose a constant M1 such that
|D1f(x)|≤M1,|D1gj(x,μ,t+αi)|≤M1, |
for (x,μ,αi)∈BׯB1(0,δ)×R. If zi∈Z1, because γi is a heteroclinic solution which is heteroclinic to the hyperbolic equilibrium p±, we can assign a constant M2 such that
|zi(t)|≤M2e−M|t|,|zi(t)+γi(t)−p±|≤M2e−M|t|. |
We define σ1(s)=f(szi(t)+γi(t))−f(γi(t)):[0,1]↦Rn. By the smoothness of f, σ1∈C3 and for some s∗∈(0,1),
f(zi(t)+γi(t))−f(γi(t))=σ1(1)−σ1(0)=σ′1(s∗)=Df(s∗zi(t)+γi(t))zi(t). |
Therefore,
|f(zi(t)+γi(t))−f(γi(t))|≤|Df(s∗zi(t)+γi(t))zi(t)|≤M1|zi(t)|≤M1M2e−M|t|. |
We define a map σ2(s):[0,1]↦Rn by
σ2(s)=gj(p±+s(zi(t)+γi(t)−p±),μ,t+αi))−gj(p±,μ,t+αi)). |
By (H4), σ2∈C3, σ2(1)=gj(γi(t)+zi(t),μ,t+αi) and σ2(0)=gj(p±,μ,t+αi)=0. For some s∗∈(0,1), we have
gj(γi(t)+zi(t),μ,t+αi)−gj(p±,μ,t+αi)=σ2(1)−σ2(0)=σ′2(s∗)=D1gj(p±+s∗(zi(t)+γi(t)−p±),μ,t+αi))(zi(t)+γi(t)−p±). |
Therefore,
|gj(γi(t)+zi(t),μ,t+αi)−gj(p±,μ,t+αi)|≤|D1gj(p±+s∗(zi(t)+γi(t)−p±),μ,t+αi))(zi(t)+γi(t)−p±)|≤M1|(zi(t)+γi(t)−p±)|≤M1M2e−M|t|. |
For any μ∈R, gj(p±,μ,t+αi)=0, thus
˜g(zi,μ,αi)(t)=˜g(zi,μ,αi)(t)−Σ2j=1μjgj(p±,μ,t+αi)=f(γi(t)+zi(t))−f(γi(t))−Df(γi(t))zi(t)+Σ2j=1μj(gj(γi(t)+z(t),μ,t+αi)−gj(p±,μ,t+αi)). |
As a result,
|˜g(zi,μ,αi)(t)|=|˜g(zi,μ,αi)(t)−Σ2j=1μjgj(p±,μ,t+αi)|≤|f(zi(t)+γi(t))−f(γi(t))|+|Df(γi(t))zi(t)|+|Σ2j=1μj(gj(γi(t)+z(t),μ,t+αi)−gj(p±,μ,t+αi))|≤(2M1M2+|μ|M1M2)e−M|t|, |
that is
||˜g(zi,μ,αi)||0=supt∈R|˜g(zi,μ,αi)(t)|eM|t|≤(2M1M2+δM1M2). |
Thus, for any given zi∈Z1, ˜g(zi,μ,αi)∈Z0. The proof is complete.
From the variable transformation of Eq (4.1), if limt→±∞|zi(t)|=0, then xi(t) is a heteroclinic solution which is heteroclinic to the hyperbolic equilibrium p− and p+. Hence, the persistence of the heteroclinic loop Γ under the periodic perturbation of Eq (1.1) is equivalent to the search solution zi(t) of Eq (4.3) in the Banach space Z1. Next, we use the method of the Lyapunov-Schmidt reduction to solve the operator equations
L1(z1)=˙z1−Df(γ1)z1=˜g(z1,μ,α1),L2(z2)=˙z2−Df(γ2)z2=˜g(z2,μ,α2), |
in the Banach space Z1.
We define spaces ˜Z1 and ˜Z2 which are closed linear subspaces of Z0, as follows
˜Z1={h∈Z0:∫∞−∞⟨φTi(t),h(t)⟩dt=0,i=1,...,d1−s},˜Z2={h∈Z0:∫∞−∞⟨ψTi(t),h(t)⟩dt=0,i=1,...,d2+s}, | (4.4) |
where φ1(t),...,φd1−s(t) are the orthonormal unit bases of Ker(L∗1) and ψ1(t),...,ψd2+s(t) are the orthonormal unit bases of Ker(L∗2). We define maps Π1 and Π2:Z0→Z0 as follows
Π1(z)(t)=d1−s∑i=1φi(t)∫∞−∞⟨φTi(t),z(t)⟩dt, | (4.5) |
Π2(z)(t)=d2+s∑i=1ψi(t)∫∞−∞⟨ψTi(t),z(t)⟩dt, | (4.6) |
where φTj and ψTj, satisfying ⟨φi,φTj⟩=δij and ⟨ψi,ψTj⟩=δij, respectively. When i=j, δij=1, and when i≠j, δij=0. By the definition of map Π1, we have
(Π1(z))2(t)=Π1(Π1(z))(t)=d1−s∑i=1φi(t)∫∞−∞⟨φTi(t),d1−s∑j=1φj(t)∫∞−∞⟨φTj(t),z(t)⟩dt⟩dt=d1−s∑i=1φi(t)d1−s∑j=1∫∞−∞⟨φTj(t),z(t)⟩dt∫∞−∞⟨φTi(t),φj(t)⟩dt=d1−s∑i=1φi(t)∫∞−∞⟨φTi(t),z(t)⟩dt=Π1(z)(t). |
For map Π2, we can similarly obtain (Π2(z))2(t)=Π2(z)(t). Hence, Π1 and Π2 are projections. For any zi∈Z1, we have
Πi(˙zi−Df(γi)zi)=0. |
Next, we apply the Lyapunov-Schmidt reduction to solve Eq (4.2). Applying Πi and (I−Πi) on Eq (4.2), we find that Eq (4.2) is equivalent to the following system
˙zi=Df(γi)zi+(I−Πi)˜g(zi,μ,αi), | (4.7) |
Πi˜g(zi,μ,αi)=0. | (4.8) |
We first solve Eq (4.7) for zi∈Z1. Then, the bifurcation equations are obtained by substituting the solution zi into Eq (4.8).
We can define a bounded linear map Ki:Ran(Li)↦Z1∖Ker(Li). Thus Ki(hi) is a solution of the linear operator equation Li(u)=˙u(t)−Df(γi)=hi, when hi∈Ran(Li). By (H6), we suppose ui1(t),...,uidi−1(t) are the orthonormal unit bases of Ker(Li). Moreover, we solve Eq (4.7) for zi∈Z1.
Lemma 2. Equation (4.7) has a unique solution zi∈Z1 such that zi satisfies
Fi(zi,βi,μ,αi)=di−1∑j=1βijuij+Ki{(I−Πi)˜g(zi,μ,αi)}, |
where (βi,μ,αi)∈Rdi−1×R2×R.
Proof. We define a C2 map: Fi:Z1×Rdi−1×R2×R→Z1 as follows:
Fi(zi,βi,μ,αi)=di−1∑j=1βijuij+Ki{(I−Πi)˜g(zi,μ,αi)}, | (4.9) |
where βi=(βi1,...,βid−1)∈Rdi−1. By Eq (4.4), we obtain ˜Zi=Ran(Li)=Ran(I−Πi), i=1,2. By the definition of the projection operator Πi and Lemma 3.1, we have (I−Πi)˜g(zi,μ,αi)∈Ran(Li). Thus Ki{(I−Πi)˜g(zi,μ,αi)} is a solution of the Eq (4.7). And uij(t)∈Ker(Li), then the fixed points of Fi are the solutions of Eq (4.7). Thus, we must demonstrate that the map Fi has a unique fixed point in the space Z1.
We let ¯B(0,δ1), ¯Bi(0,δ2), and ¯B1(0,δ2) be a closed subset with radius δ1>0 and δ2>0 centered at the origins of Z1, Rdi−1, and R2. By ˜g(0,0,αi)=0 and the smoothness of f,gj, we can set δ1 and δ2 to be sufficiently small such that
||˜g(zi,μ,αi)||0<δ2, |
for (zi,μ,αi)∈¯B(0,δ1)ׯB1(0,δ2)×R.
Further, uij∈Ker(Li), Ki and (I−Πi) are bounded linear operators. We can set constants M3>0,M4>0 such that
||uij||1≤M3,||Ki(I−Πi)||≤M4, |
for any i=1,2,j=1...,di−1. We let δ2=min{δ12M3(di−1),δ12M4}. For any (zi,βi,μ,αi)∈ׯB(0,δ1)ׯBi(0,δ2)ׯB1(0,δ2)×R, we have
||Fi(zi,βi,μ,αi)||1=||di−1∑j=1βijuij+Ki{(I−Πi)˜g(zi,μ,αi)||1≤||di−1∑j=1βijuij||1+||Ki{(I−Πi)˜g(zi,μ,αi)||1≤δ2(di−1)M3+δ2M4≤δ1. |
Thus, for any (βi,μ,αi)∈¯Bi(0,δ2)ׯB1(0,δ2)×R, we have
Fi(⋅,βi,μ,αi):¯B(0,δ1)↦¯B(0,δ1). |
We let
h(zi)(t)=f(γi(t)+zi(t))−f(γi(t))−Df(γi(t))zi(t). |
Then h(0)=0,Dh(0)=0, so we can choose above δ1 to be sufficiently small such that ||Dh(zi)||≤δ2, for zi∈¯B(0,δ1). We select a constant M5>0 such that ||D1gj(γi(t)+zi(t),μ,t+αi)||≤M5, for (zi,μ,αi)∈ׯB(0,δ1)ׯB1(0,δ2)×R.
By Eq (4.3), we have
˜g(zi,μ,αi)(t)=h(zi)(t)+Σ2j=1μj(gj(γi(t)+zi(t),μ,t+αi). |
For z1i,z2i∈¯B(0,δ1), (βi,μ,αi)∈¯Bi(0,δ2)ׯB1(0,δ2)×R. From Eq (4.3), we obtain the following:
||Fi(z1i,βi,μ,αi)−Fi(z2i,βi,μ,αi)||=||Ki{(I−Πi){˜g(z1i,μ,αi)}−Ki{(I−Πi)˜g(z2i,μ,αi)}||=||Ki{(I−Πi){˜g(z1i,μ,αi)−˜g(z2i,μ,αi)}||=||Ki{(I−Πi){h(z1i(t))−h(z2i(t))+Σ2j=1μj(gj(γi(t)+z1i(t),μ,t+αi)−gj(γi(t)+z2i(t),μ,t+αi))||≤||Ki(I−Πi)||{|Dh(z1i(t)+s(z2i(t)−z1i(t)))||z1i(t)−z2i(t)|+Σ2j=1|μj||(gj(γi(t)+z1i(t),μ,t+αi)−gj(γi(t)+z2i(t),μ,t+αi))|}≤||Ki(I−Πi)||{Dh(z1i(t)+s(z2i(t)−z1i(t)))+Σ2j=1|μj|(D1gj(γi(t)+z1i(t)+s(z1i(t)−z2i(t)),μ,t+αi)}||z1i−z2i||, |
for s∈(0,1). Thus,
||Fi(z1i,βi,μ,αi)−Fi(z2i,βi,μ,αi)||≤δ2(M4+2M5)||z1i−z2i||. |
Therefore, if we set δ2=min{δ12M3(di−1),δ12M4,12(M4+2M5)}, then
||Fi(z1i,βi,μ,αi)−Fi(z2i,βi,μ,αi)||≤12||z1i−z2i||. |
As a result, Fi is a uniform contraction in ¯B(0,δ1). By the contraction mapping principle, a unique C1 map ωi:¯Bi(0,δ)ׯB1(0,δ)×R↦Z1 exists such that
ωi(βi,μ,αi)=di−1∑j=1βijuij+Ki{(I−Πi)˜g(ωi,μ,αi)}. |
Moreover, Fi(0,0,0,αi)=0, hence, ωi(0,0,αi)=0, which implies the desired statement.
Substituting ωi(βi,μ,αi) into Eq (4.8), we obtain the bifurcation function
0=Πi˜g(ωi(βi,μ,αi),μ,αi). | (4.10) |
By the definition of projection Πi, we have
d1−s∑i=1ψi(t)∫+∞−∞⟨ψTi(s),˜g(ω1(β1,μ,α1),μ,α1)(s)⟩ds=0, | (4.11) |
d2+s∑i=1φi(t)∫+∞−∞⟨φTi(s),˜g(ω2(β2,μ,α2),μ,α2)(s)⟩ds=0. | (4.12) |
By the linear independence of φ1,...,φd1−s and ψ1,...,ψd2+s(t), Eqs (4.11) and (4.12) are equivalent to
H1i(β1,μ,α1)=∫+∞−∞⟨ψTi(s),˜g(ω1(β1,μ,α1),μ,α1)(s)⟩ds=0, | (4.13) |
H2j(β2,μ,α2)=∫+∞−∞⟨φTj(s),˜g(ω2(β2,μ,α2),μ,α2)(s)⟩ds=0, | (4.14) |
where i=1,...d1−s,j=1,...,d2+s. We let
H1(β1,μ,α1)=(H11(β1,μ,α1),...,H1d1−s(β1,μ,α1)),H2(β2,μ,α2)=(H21(β2,μ,α2),...,H2d2+s(β2,μ,α2)). |
Therefore, by the Lyapunov-Schmidt reduction, we obtained the bifurcation function:
H(β,μ,α)=(H1(β1,μ,α1),H2(β2,μ,α2)), |
where β=(β1,β2),α=(α1,α2). If there are some parameter values (β,μ,α)∈Rd1+d2−2×R2×R2, such that
H(β,μ,α)=0, |
then zi=ωi is a solution of Eq (4.2). Hence, the perturbed Eq (1.2) has heteroclinic solutions
x1(β1,μ,α1)(t)=γ1(t)+ω1(β1,μ,α1)(t), |
and
x2(β2,μ,α2)(t)=γ2(t)+ω2(β2,μ,α2)(t), |
which are asymptotic to the equilibrium p+ and p−, that is
limt→+∞x1(β1,μ,α1)(t)=p+,limt→−∞x1(β1,μ,α1)(t)=p− |
and
limt→+∞x2(β2,μ,α1)(t)=p−,limt→−∞x2(β2,μ,α1)(t)=p+, |
are uniform for some (β1,β2,μ,α1,α2). Thus, the heteroclinic orbits x1(β1,μ,α1)(t) and x2(β2,μ,α2)(t) and the equilibria p+, p− constitute a heteroclinic loop of the perturbed Eq (1.2).
Through direct calculations, the function H(β,μ,α) has the following properties:
(i)H1(0,0,α1)=H2(0,0,α2)=0,∂H1i∂β1p(0,0,α1)=∂H2j∂β2q(0,0,α2)=0;(ii)∂2H1i∂β1p∂β1q(0,0,α1)=∫+∞−∞⟨ψTi(s),D11f(γ1(s))u1p(s)u1q(s)⟩ds;(iii)∂2H2j∂β2p∂β2q(0,0,α2)=∫+∞−∞⟨φTi(s),D11f(γ2(s))u2p(s)u2q(s)⟩ds;(iv)∂H1i∂μk(0,0,α1)=∫+∞−∞⟨ψTi(s),gk(γ1(s),μ,s+α1)⟩ds;(v)∂H2j∂μk(0,0,α2)=∫+∞−∞⟨φTi(s),gk(γ2(s),μ,s+α2)⟩dt. |
We define M1:Rd1−1×R2×R→Rd1−s given by
M1(β1,μ,α1)=(M11(β1,μ,α1),...,M1d1−s(β1,μ,α1)), |
and
M1i(β1,μ,α1)=2∑k=1a1i,k(α1)μk+12d1−1∑p=1d1−1∑q=1b1i,pqβ1pβ1q,i=1,...,d1−s. |
We define M2:Rd2−1×R2×R→Rd2+s given by
M2(β2,μ,α2)=(M21(β2,μ,α2),...,M2d2+s(β2,μ,α2)), |
and
M2j(β2,μ,α2)=2∑k=1a2j,k(α2)μk+12d2−1∑p=1d2−1∑q=1b2j,pqβ2pβ2q,j=1,...,d2+s. |
Thus,
Hi(βi,μ,αi)=Mi(βi,μ,αi)+H.O.T. |
Moreover, we define M:Rd1+d2−2×R2×R2→Rd1−s×Rd2+s given by
M(β,μ,α)=(M1(β1,μ,α1),M2(β2,μ,α2)), |
hence
H(β,μ,α)=M(β,μ,α)+H.O.T. |
Lemma 3. If points (β0,μ0,α0)∈Rd1+d2−2×R2×R2 exists such that M(β0,μ0,α0)=0, and D(β,μ)M(β0,μ0,α0) is a nonsingular (d1+d2)×(d1+d2) matrix, then an open interval I⊂R exists containing zero and differentiable functions, κ1:I→Rd1+d2−2 and κ2:I→R2, such that κ1(0)=0, κ2(0)=0, and H(ε(β0+κ1(ε)),ε2(μ0+κ2(ε)),α0)=0 for ε∈I.
Proof. We define a C2 function N:Rd1+d2−2×R2×R2↦Rd1+d2:
N(x,y,ε)={1ε2H(ε(β0+x),ε2(μ0+y),α0),forε≠0,M(β0+x,μ0+y,α0),for ε=0. |
It is clear that H=0 if and only if N=0 for ε≠0. Through direct calculations, we have N(0,0,0)=0, and D(x,y)N(0,0,0)=D(β,μ)M(β0,μ0,α0) is nonsingular matrix. Using the implicit function theorem, we know an open interval I⊂R exists containing the zero and differentiable functions, which are κ1:I→Rd1+d2−2 and κ2:I→R2, satisfying κ1(0)=0 and κ2(0)=0, respectively, such that N(κ1(ε),κ2(ε),ε)=0 for ε∈I. Hence, we obtain
H(ε(β0+κ1(ε)),ε2(μ0+κ2(ε)),α0)=0forε∈I∖{0}. |
The proof is complete.
Hence, the perturbed Eq (1.2) has heteroclinic orbits
and
where . In addition,
for some . If we let
then some solutions near the unperturbed heteroclinic loop exist which can constitute a heteroclinic loop for perturbed Eq (1.2).
In this paper, we investigated the persistence of a heteroclinic loop under periodic perturbation in . We assumed unperturbed heteroclinic loop is a heterodimensional loop and the unperturbed heteroclinic orbits are degenerate. Using the method of Lyapunov-Schmidt reduction and exponential dichotomies, we obtained the bifurcation function, which is defined by
where and . Under the condition of Theorem 1, there exist some points such that . Hence, there exist heteroclinic solutions , of the Eq (1.2) with , where , and are located near the heteroclinic orbits and , such that , , and can constitute a heteroclinic loop . The heteroclinic tangles is one of the primary mechanisms for non-uniformly hyperbolic dynamics. Our results extended the theory of heteroclinic loop bifurcation.
There are still many interesting and instructive issues worthy of further study. For example, the hyperbolicity of the heteroclinic solution and chaos motion near the heteroclinic loop .
We are grateful to the two anonymous referees for useful comments and suggestions. This work was supported by National Natural Science Foundation of China (Grant No. 11801343).
The authors declare there is no conflicts of interest.
[1] | J. R. Guckenheimer, P. Holmes, Nonlinear Oscillations, Dynamical Systems, and Bifurcations of Vector Fields, Springer-Verlag, New York, 1983. |
[2] |
K. H. Alfsen, J. Fryland, Systematics of the Lorenz model at , Phys. Scr., 31 (1985), 15–20. https://doi.org/10.1088/0031-8949/31/1/003 doi: 10.1088/0031-8949/31/1/003
![]() |
[3] | C. Bonatti, L. J. Díaz, M. Viana, Dynamics Beyond Uniform Hyperbolicity: A Global Geometric and Probabilistic Perspective, Springer-Verlag, New York, 2005. |
[4] | M. Han, J. Yang, A. Tarta, Y. Gao, Limit cycles near homoclinic and heteroclinic loops, J. Dyn. Differ. Equations, 20 (2008). https://doi.org/10.1007/s10884-008-9108-3 |
[5] |
X. Sun, M. Han, J. Yang, Bifurcation of limit cycles from a heteroclinic loop with a cusp, Nonlinear Anal. Theory Methods Appl., 74 (2011), 2948–2965. https://doi.org/10.1016/j.na.2011.01.013 doi: 10.1016/j.na.2011.01.013
![]() |
[6] |
F. Chen, A. Oksasoglu, Q. Wang, Heteroclinic tangles in time-periodic equations, J. Differ. Equations, 254 (2013), 1137–1171. https://doi.org/10.1016/j.jde.2012.10.010 doi: 10.1016/j.jde.2012.10.010
![]() |
[7] | I. S. Labouriau, A. A. P. Rodrigues, Periodic forcing of a heteroclinic network, J. Dyn. Differ. Equations, 2021 (2021). https://doi.org/10.1007/s10884-021-10054-w |
[8] |
I. S. Labouriau, A. A. P. Rodrigues, Bifurcations from an attracting heteroclinic cycle under periodic forcing, J. Differ. Equations, 269 (2020), 4137–4174. https://doi.org/10.1016/j.jde.2020.03.024 doi: 10.1016/j.jde.2020.03.024
![]() |
[9] |
S. N. Chow, B. Deng, D. Terman, The bifurcation of homoclinic and periodic orbits from two heteroclinic orbits, SIAM J. Math. Anal., 21 (1990), 179–204. https://doi.org/10.1137/0521010 doi: 10.1137/0521010
![]() |
[10] |
D. Zhu, Z. Xia, Bifurcations of heteroclinic loops, Sci. China Ser. A Math., 41 (1998), 837–848. https://doi.org/10.1007/BF02871667 doi: 10.1007/BF02871667
![]() |
[11] |
J. D. M. Rademacher, Homoclinic orbits near heteroclinic cycles with one equilibrium and one periodic orbit, J. Differ. Equations, 218 (2005), 390–443. https://doi.org/10.1016/j.jde.2005.03.016 doi: 10.1016/j.jde.2005.03.016
![]() |
[12] |
X. Lin, Using Melnikov's method to solve Shilnikov's problems, Proc. R. Soc. Edinburgh Sect. A Math., 116 (1990), 295–325. https://doi.org/10.1017/S0308210500031528 doi: 10.1017/S0308210500031528
![]() |
[13] |
F. Geng, T. Wang, X. Liu, Global bifurcations near a degenerate hetero-dimernsional cycle, J. Appl. Anal. Comput., 8 (2018), 123–151. https://doi.org/10.11948/2018.123 doi: 10.11948/2018.123
![]() |
[14] |
X. Liu, X. Wang, T. Wang, Nongeneric bifurcations near a nontransversal heterodimensional cycle, Chin. Ann. Math. Ser. B, 39 (2018), 111–128. https://doi.org/10.1007/s11401-018-1055-7 doi: 10.1007/s11401-018-1055-7
![]() |
[15] |
A. Vanderbauwhede, Bifurcation of degenerate homoclinics, Results Math., 21 (1992), 211–223. https://doi.org/10.1007/BF03323080 doi: 10.1007/BF03323080
![]() |
[16] | S. N. Chow, J. K. Hale, Methods of Bifurcation Theory, Springer-Verlag, New York, 1982. |
[17] | X. Lin, Lin's method, Scholarpedia, 3 (2008), 6972. |
[18] |
S. N. Chow, J. K. Hale, J. Mallet-Parret, An example of bifurcation to homoclinic orbits, J. Differ. Equations, 37 (1980), 551–573. https://doi.org/10.1016/0022-0396(80)90104-7 doi: 10.1016/0022-0396(80)90104-7
![]() |
[19] |
K. J. Palmer, Transversal heteroclinic points and Cherry's example of a nonintegrable Hamiltonian system, J. Differ. Equations, 65 (1986), 321–360. 10.1016/0022-0396(86)90023-9 doi: 10.1016/0022-0396(86)90023-9
![]() |
[20] |
J. K. Hale, X. Lin, Heteroclinic orbits for retarded functional differential equations, J. Differ. Equations, 65 (1986), 175–202. https://doi.org/10.1016/0022-0396(86)90032-X doi: 10.1016/0022-0396(86)90032-X
![]() |
[21] |
J. R. Gruendler, Homoclinic solutions for autonomous systems in arbitrary dimension, SIAM J. Math. Anal., 23 (1992), 702–721. https://doi.org/10.1137/0523036 doi: 10.1137/0523036
![]() |
[22] |
J. Gruendler, Homoclinic solutions for autonomous ordinary differential equations with nonautonomous perturbation, J. Differ. Equations, 122 (1995), 1–26. https://doi.org/10.1006/jdeq.1995.1136 doi: 10.1006/jdeq.1995.1136
![]() |
[23] |
R. J. Sacker, The splitting index for linear differential systems, J. Differ. Equations, 33 (1979), 368–405. https://doi.org/10.1016/0022-0396(79)90072-X doi: 10.1016/0022-0396(79)90072-X
![]() |
1. | Bin Long, Yiying Yang, Applying Lin's method to constructing heteroclinic orbits near the heteroclinic chain, 2024, 47, 0170-4214, 10235, 10.1002/mma.10118 | |
2. | Bin Long, Yiying Yang, Periodic or homoclinic orbit bifurcated from a heteroclinic loop for high-dimensional systems, 2025, 23, 2391-5455, 10.1515/math-2025-0134 |