Abstract
We are concerned with the existence of positive periodic solutions of third-order periodic boundary value problems
where \(k_1,k_3\in (0,\infty )\) and \(k_2\in (0,(\frac{\pi }{\omega })^2)\) are constants. \(\lambda \) is a positive parameter. The weight function \(a\in C([0,\omega ],\mathbb {R})\) may change sign. \(f\in C([0,\infty ),\mathbb {R})\) with \(f(0):=\lim \nolimits _{u\rightarrow 0^+}f(u)>0\). We show that there exists a constant \(\lambda ^*>0\), such that the above problem has at least one positive periodic solution for \(\lambda \in (0,\lambda ^*)\) where f(0) is bounded. This result is based upon bifurcation theory and Leray–Schauder fixed point theorem. On the other hand, by using Krasnoselskii’s fixed point theorem in a cone, we show that there exists a positive constant \(\lambda _0\) such that for all \(\lambda \in (0,\lambda _0)\), the above problem has at least one positive periodic solution where f(0) is unbounded, namely that f has a singularity at \(u=0\). And this result is applicable to weak as well as strong singularities.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
1 Introduction
In this paper, we proceed with the investigation of positive solutions for a class of third-order differential equations
subjects to the periodic boundary conditions
where \(k_1,k_3\in (0,\infty )\) and \(k_2\in (0,(\frac{\pi }{\omega })^2)\) are constants and \(\lambda \) is a positive parameter. \(a:[0,\omega ]\rightarrow \mathbb {R}\) is a continuous function which may change sign. Define that \(a^{+}(t):=\max \nolimits _{t\in [0,\omega ]}\{a(t),0\}\) and \(a^{-}(t):=\max \nolimits _{t\in [0,\omega ]}\{-a(t),0\}\). \(f:[0,\infty )\rightarrow \mathbb {R}\) is a continuous function with \(f(0):=\lim \nolimits _{u\rightarrow 0^+}f(u)>0\).
For the nonlinear third-order periodic boundary value problems, we recall the following results. In [1], by using Schauder fixed point theorem, together with perturbation technique, Kong and his collaborators established the existence of positive periodic solutions for (1.1)–(1.2) with \(k_1=k_2=0\), \(k_3\in (0,\frac{1}{\sqrt{3}}))\) and \(\lambda =1\), where a(t)f(u) is a nonnegative function defined on \([0,\omega ]\times (0,\infty )\) and is nonincreasing in \(u>0\) for \(t\in [0,\omega ]\) and
Typical example of nonlinearity in [1] is of the form \(f(u)=\frac{1}{u^{\nu }}\). Afterwards, Sun [2] and Chu [3] used different methods about fixed point theorems to generalize the results of [1] under the a(t)f(u) satisfies suitable conditions.
In [4], Feng proved the existence of positive periodic solutions of (1.1)–(1.2) with \(k_1>0\), \(k_2>0\), \(k_3=0\) and \(\lambda =1\), by using the well-known Guo-Krasnoselskii’s fixed point theorem where a(t)f(u) is nonnegative. Recently, Ren and her collaborators [5, 6] made an exhaustive study of Green’s function of the third-order linear differential equation \(u'''(t)+\breve{a} u''(t)+\breve{b} u'(t)+\breve{c}u(t)=\breve{h}(t)\). Also, by applying the properties of the Green’s function, the authors showed that the existence of positive periodic solutions corresponding to third-order singular differential equation
where \(\breve{f}\in \text{ Car }~(\mathbb {R}\times (0,+\infty ),(0,+\infty ))\) is nonnegative and has a singularity at \(u=0\), \(\breve{e}\in L^{1}(\mathbb {R})\) is an \(\omega \)-periodic function.
Notice that all the aforementioned results in these works use a key condition that the nonlinear term is nonnegative, that is, a is nonnegative. The terminology “indefinite” was probably introduced by Hess and Kato [7] under the framework of linear eigenvalue problems. In a recent series of papers, it is worth mentioning that the qualitative study of the solutions to the indefinite equations has been treated by many authors in both ODEs and PDEs. For example, by using Schauder fixed point theorem, Hai [8] showed the existence of positive solutions of the semilinear elliptic boundary value problem with \(\hat{a}\) changing-sign
And the study the existence of positive periodic solutions for singular differential equations with an indefinite weight began from 2010, Bravo and Torres [9] discussed a special second-order differential equation (Emden–Flower equation) with \(\hat{b}\) changing-sign
Recently, the study of existence of positive periodic solutions for second-order singular differential equations has attracted by many researchers’ attention, see [10,11,12,13].
Motivated by the above works and focus on the indefinite weight, our main results are the following.
Theorem 1.1
Assume that \((A_1)-(A_2)\) are satisfied.
\((A_1)\) If f(0) is bounded, then there exists two constants \(\epsilon \) and \(\delta \in (0,1)\) such that \(f(x)\ge \delta f(0)\) for \(x\in (0,\epsilon )\).
\((A_2)\) \(a\in C([0,\omega ],\mathbb {R})\), \(a\not \equiv 0\), and there exists a number \(k>1\) such that
Then there exists a positive constant \(\lambda ^{*}>0\) such that (1.1)–(1.2) has at least one positive periodic solution for \(\lambda \in (0,\lambda ^{*})\).
Theorem 1.2
If f(0) is unbounded, that is, f has a singularity at \(u=0\) in the form of \(f(u):=\frac{1}{u^{\nu }}\) where \(\nu \in (0,\infty )\) is a constant. Then suppose that \(1-\sigma ^*\omega >0\) and
are satisfied. Then there exists a positive constant \(\lambda _0\) such that (1.1)–(1.2) admits at least one positive periodic solution for \(\lambda \in (0,\lambda _0)\).
Remark 1.3
In the case that f(0) is bounded, condition \((A_2)\) covers the case that a(t) is positive for all \(t\in [0,\omega ]\). In this paper, we show the global structure of positive periodic solutions for (1.1)–(1.2) with the positive weight by using bifurcation theory and we mention that our result is new in the case the weight function is positive.
Remark 1.4
In the case that f(0) is unbounded, that is, this problem can be regarded as a class of indefinite singular problems.
2 Preliminary Results
Lemma 2.1
([15, Lemma 2.4]) Let \(\tilde{E}\) be a Banach space, and let \(\{C_{n}\}\) be a family of closed connected subsets of \(\tilde{E}\). Assume that
-
(i)
there exists \(z_{n}\in C_{n}\), \(n=1,2,\ldots \), and \(z_{*}\in \tilde{E}\) such that \(z_{n}\rightarrow z_{*}\);
-
(ii)
\(\lim \nolimits _{n\rightarrow \infty }r_{n}=\infty \), where \({r_{n}=\sup \{\Vert x\Vert _{\infty }|x\in C_{n}\}}\);
-
(iii)
for every \(R>0\), \((\bigcup ^{\infty }_{n=1}C_{n})\cap B_{R}\) is a relatively compact set of \(\tilde{E}\), where
$$\begin{aligned} B_{R}=\{x\in \tilde{E}|~\Vert x\Vert _{\infty }\le R\}. \end{aligned}$$Then \(\mathcal {D}:=\limsup \nolimits _{n\rightarrow +\infty } C_{n}\) contains an unbounded component \(\mathcal {C}\) with \(z_{*}\in \mathcal {C}\).
Lemma 2.2
Assume that \((A_1)\) is satisfied and \(\delta \in (0,1)\). Then there exists a number \(\bar{\lambda }>0\), such that (1.1)–(1.2) has a positive periodic solution \(u_{\lambda }\) for all \(\lambda \in (0,\bar{\lambda })\) with
To use bifurcation theorem to prove Lemma 2.2, we extend f to
Let \(X=C[0,\omega ]\) be the Banach space with its usual normal \(\Vert u\Vert _{\infty }:=\max \limits _{u\in [0,\omega ]}|u(t)|\). Let us consider
Define a linear operator L: \(E\rightarrow X\)
with the Domain
Next, we show that the conditions of \(k_1,~k_2,~k_3\) of inverse positive operator L.
Lemma 2.3
If (2.1) has nontrivial solutions, then \(k_1,k_3\in (0,\infty )\) and \(k_2\in (0,(\frac{\pi }{\omega })^{2})\).
Proof
Let y be a nontrivial solution of (2.1) and
for some \(k_1,~k_2\in \mathbb {R}\) and \(k_3=k_1k_2\).
If \(y''+k_2y=0\), \(y(0)=y(\omega ),~y'(0)-y'(\omega )=1\), then we have \(k_2\in (0,(\frac{\pi }{\omega })^2)\) and consequently
In this case we obtain that \(k_1,k_3\in \mathbb {R}\) and \(k_2\in (0,\left( \frac{\pi }{\omega }\right) ^2)\).
If \(y''+k_2y\ne 0\), then
This implies that \(y'+k_1y=0\), \(y(0)-y(\omega )=1\), i.e., \(k_1\in (0,\infty )\) and
A simple computation yields
where \(k_2\in (0,\left( \frac{\pi }{\omega }\right) ^2)\) from [14, Lemma 3.2]. Since \(k_3=k_1k_2\), then \(k_3\in (0,\infty )\). \(\square \)
It is worth mentioning that if L is inverse positive, then L is invertible on E and the Green’s function G(t, s) related to the problem
is positive, here \(h\in X\) and \(h\not \equiv 0\).
To do that, let us define a linear operator
Then the Green function of \(L_1y=\hat{h}_1(t)\) is
Define a linear operator
Then the Green’s function of \(L_2y=\hat{h}_2(t)\) is
Obviously,
then Green’s function of \(Ly=h(t)\) is
i.e.,
Lemma 2.4
Let \(k_2\in (0,(\frac{\pi }{\omega })^2)\) and \(k_1,k_3\in (0,\infty )\). Then
Proof
It is an immediate consequence of the fact that for \(j=1,2\),
\(\square \)
Lemma 2.5
(Maximum principle) Let \(k_2\in (0,(\frac{\pi }{\omega })^2)\), \(k_1,k_3\in (0,\infty )\) and \(L(\tilde{y}(t))\ge 0\) for \(\tilde{y}\in E\), then \(\tilde{y}(t)\ge 0\) on \([0,\omega ]\).
Proof
Let \(\tilde{y}(t)\in E\) satisfies \(L(\tilde{y}(t))\ge 0\) on \([0,\omega ]\). The \(\tilde{y}\) is a solution of BVP \(L(\tilde{y}(t))=\tilde{h}(t)\) and \(\tilde{h}(t)\) is a nonnegative continuous function. Thus the function \(\tilde{y}\) is a solution of an integral equation
since \(G(t,s)>0\) on \([0,\omega ]\times [0,\omega ]\). \(\square \)
So (2.1) is equivalent to the integral equation
Let P be the cone in X defined by
where \(\sigma :=\frac{\sigma _*}{\sigma ^*}\), \(\sigma ^*=\max \nolimits _{0\le t,s\le \omega }G(t,s)\), \(\sigma _*=\min \nolimits _{0\le t,s\le \omega }G(t,s)\) and \(A:P\rightarrow P\) be the map defined by
Then from condition \((A_1)\) and the positivity of G(t, s), \(A(P)\subset P\) and \(A:P\rightarrow P\) is compact and continuous. So the existence of positive solutions for (2.1) is equivalent to the existence of a fixed point of A in P.
Next, we consider the principal eigenvalue of the linear problem
Lemma 2.6
Suppose that \(a\in C([0,\omega ], \mathbb {R})\) with \(a(t)\not \equiv 0\) on any subinterval of \([0,\omega ]\), and \(a^{+}(t)>0\) for all \(t\in [0,\omega ]\). Then (2.6) has a principal eigenvalue \(\lambda _{1}\), which is positive and simple, and the corresponding eigenfunction \(\varphi _1(t)\) is positive on \([0,\omega ]\).
Proof
From the definition of cone P, we know that P is normal and has nonempty interior. Obviously, \(X=P-P\). Since
it follows that the linear operator
is a strong positive operator, that is, \(\bar{L}u(t)\in \text{ int }~P\). By Krein–Rutman Theorem [16], Theorem 19.3(a)], the spectral radius \(r(\bar{L})\) is positive, and there exists a function \(\varphi _1\in E\) such that \(\varphi _1>0\) on \([0,\omega ]\) and \(\bar{L}\varphi _1=r(\bar{L})\varphi _1\). Thus, \(\lambda _1=(r(\bar{L}))^{-1}>0\). Let \(\bar{L}^*\) is conjugate operator of \(\bar{L}\), and \(\bar{L}^*\psi _1=\lambda _1\psi _1\), where \(\psi _1\in E\) such that \(\psi _1>0\) on \([0,\omega ]\) corresponding to \(\lambda _1\). Since
where \(\langle \cdot ,\cdot \rangle \) denotes the standard \(L^2(0,\omega )\) inner product, then the eigenvalue \(\lambda _1\) has multiplicity 1. Thus, \(\lambda _1=(r(\bar{L}))^{-1}\) is the principal eigenvalue of (2.6). \(\square \)
Lemma 2.7
Assume that \((A_1)\) holds. If \(u\in \partial \Omega _r\), \(r>0\) and \(\Omega _r:=\{t\in [0,\omega ]|~\Vert u\Vert _{\infty }<r\}\), then
where \(\widehat{M_r}=1+\max _{0\le s\le r}\{\tilde{f}(s)\}\).
Proof
Since \(\tilde{f}(u(t))\le \widehat{M_r}\) for \(t\in [0,\omega ]\), it follows that
\(\square \)
Lemma 2.8
Assume that \(\{(\mu _k,y_{k})\}\subset (0,\infty )\times P\) is a sequence of positive solutions of (2.1). Suppose that \(\mu _k<C_0\) for some constant \(C_0>0\) and
Then
Proof
Suppose by the way of contradiction that \(\{\Vert y_{k}\Vert _{\infty }\}\) is bounded. It follows that
where \(\beta \) is independent of k and is a constant. Thus,
that \(\{y'''_k\}\) is uniformly bounded in C[0, 1], and subsequently \(\{y''_k\}\) is uniformly bounded in C[0, 1], which contradicts (2.8). \(\square \)
Proof of Lemma 2.2
Let \(\Sigma \) be the closure of the set of positive periodic solutions of (2.1) in E. To prove Lemma 2.2, we use bifurcation approach to treat the case \(\tilde{f}_{0}:=\lim \nolimits _{u\rightarrow 0^+}\frac{\tilde{f}(u)}{u}=\infty \). In this approach, it is crucial to construct a sequence of function \(\{\tilde{f}^{[n]}\}\) such that \(\{\tilde{f}^{[n]}\}\) is asymptotic linear at 0 and satisfies
From corresponding auxiliary equations, we obtain a sequence of unbounded components \(\{C^{[n]}\}\) via nonlinear Krein–Rutman bifurcation theorem, see Dancer [17] and Zeidler [18]. So we can find unbounded components \(\mathcal {C}\) which satisfies
and joins (0, 0) with \((0,\infty )\).
To apply the bifurcation theorem, we extend \(\tilde{f}\) to a function \(\{g^{[n]}\}:\mathbb {R}\rightarrow \mathbb {R}\) by
then
Thus,
Let us consider the auxiliary family of the equations
Let \(\zeta \in C(\mathbb {R})\) be such that
then
Let us consider
as a bifurcation problem from the trivial solution \(u\equiv 0\).
(2.12) can be converted to the equivalent the equation
Further we note that \(\Vert L^{-1}[a^{+}(\cdot )\zeta ^{[n]}(u(\cdot ))]\Vert _{\infty }=o(\Vert u\Vert _{\infty })\) for u near 0 in E.
From the fact \((g^{[n]})_0>0\), the results of nonlinear Krein–Rutman theorem for (2.12) can be stated as follows: there exists a continuum \(C^{[n]}\) of positive solutions of (2.12) joining \((\frac{\lambda _1}{(g^{[n]})_0})\) to infinity in P. Moreover, \(C^{[n]}\backslash \{(\frac{\lambda _1}{(g^{[n]})_0},0)\}\subset \text{ int }~P\) and \((\frac{\lambda _1}{(g^{[n]})_0},0)\) is the only positive bifurcation point of (2.12) lying on trivial solutions line \(u\equiv 0\).
Next, we show that \(\{C^{[n]}\}\) satisfies all of the conditions of Lemma 2.1. Since
condition (i) in Lemma 2.1 is satisfied with \(z^*=(0,0)\). Obviously
so (ii) holds. From the Arzela–Ascoli Theorem and the definition of \(g^{[n]}\), we can get that (iii) is satisfied. Therefore, the superior limit of \(\{C^{[n]}\}\), that is, \(\mathcal {D}\) contains an unbounded connected component \(\mathcal {C}\) with \((0,0)\in \mathcal {C}\).
Let \(\{\mu _{k},y_{k}\}\subset \mathcal {C}\) be such that \(|\mu _{k}|+\Vert y_{k}\Vert _{\infty }\rightarrow \infty \) as \(k\rightarrow \infty \). Then
We divide the proof into two steps.
Step 1. We show that \(\sup \{\Vert u\Vert _{\infty }|(\lambda ,u)\in C^{[n]}\}=\infty \).
Suppose by the way of contradiction that \(\sup \{\Vert u\Vert _{\infty }|(\lambda ,u)\in C^{[n]}\}=:\beta _1<\infty \). Since \(\{\mu _{k},y_{k}\}\subset \mathcal {C}\) is such that \(|\mu _{k}|+\Vert y_{k}\Vert _{\infty }\rightarrow \infty \), then we have
Note that
Since \(a^{+}(t)>0\) for \(t\in [a,b]\subset [0,\omega ]\). So, there exists a constant \(\beta _2>0\), such that
Combining (2.15) and (2.16) with the relation
Thus, from (2.17) and the remarks in the final paragraph on P. 56 of [19], we deduce that \(y_k\) must change its sign on [a, b] if k is large enough. This is a contradiction. Hence \(\{\Vert y_{k}\Vert _{\infty }\}\) is unbounded.
Step 2. We show that \(\sup \{\lambda |(\lambda ,u)\in C^{[n]}\}=\infty \).
Now, taking \(\{\mu _{k},y_{k}\}\subset \mathcal {C}\) be such that
We show that \(\lim \nolimits _{k\rightarrow \infty }\mu _{k}=0.\) Suppose on the contrary that, choosing a subsequence and relabelling if necessary, \(\mu _k\ge \mu _1\) for some constant \(\mu _1>0\). Then from (2.18) and \((A_2)\), we have
Thus, from (2.17) and from the remarks in the final paragraph on P. 56 of [19], we deduce that \(y_k\) must change its sign on [a, b] if k is large enough. This is a contradiction. Therefore, \(\lim \nolimits _{k\rightarrow \infty }\mu _k=0\).
We rewrite (2.10) to
By Lemma 2.7, for every \(r>0\) and \(u\in \partial \Omega _r\),
where \(\widehat{M_r}=1+\max _{0\le s\le r}\{g^{[n]}(u(s))\}\).
Let there exists a constant \(\bar{\lambda }\) be such that
Then for \(\lambda \in (0,\bar{\lambda })\) and \(u\in \partial \Omega _r\), \(\Vert Lu\Vert _{\infty }<\Vert u\Vert _{\infty }\). This mean that
i.e., \(\mathcal {C}\) is also an unbounded component joining (0, 0) with \((0,\infty )\) in \([0,\infty )\times X\). Thus, for \(\lambda \in (0,\bar{\lambda })\), (2.1) has at least one positive solution. Moreover, from \((A_1)\) we have \(u_{\lambda }\ge \lambda \delta f(0)\int ^\omega _0G(t,s)a^{+}(s)ds\). \(\square \)
3 Proof of Theorem 1.1
From \((A_2)\), there exist two positive constants \(\alpha ,\gamma \in (0,1)\) such that
for \(x\in [0,\alpha ]\), \(t\in [0,1]\). Fixed \(\delta \in (\gamma ,1)\) and let \(\lambda ^*>0\) be such that
for \(\lambda <\lambda ^*\), where \(\tilde{u}_{\lambda }\) is given by Lemma 2.2, from \((A_1)\),
and
for \(x,y\in [-\alpha ,\alpha ]\) with \(|x-y|\le \lambda ^*\delta f(0)\int ^\omega _0G(t,s)a^{+}(s)ds\).
Next, note that \(\lambda <\lambda ^*\). We find a solution \(u_{\lambda }\) of (1.1)–(1.2) has the form \(\tilde{u}_{\lambda }+v_{\lambda }\). Thus \(v_{\lambda }\) solves
For all of \(w\in C([0,\omega ])\), we suppose that \(v=Aw\) is the solution of
Then A is completely continuous. Let \(v\in C([0,\omega ])\) and \(\theta \in (0,1)\) be such that \(v=\theta Av\). Then
Next, We show that \(\Vert v\Vert _{\infty }\ne \lambda \delta f(0)\int ^\omega _0G(t,s)a^{+}(s)ds\). Suppose not. Then, from (3.1) and (3.2), it follows that
and
which together with (3.1) implies that
which is a contradiction. From Leray–Schauder fixed point theorem, A has a fixed point
Hence \(v_{\lambda }\) satisfies (3.4) and
i.e., \(u_{\lambda }\) is a positive solution of (1.1)–(1.2). This completes the proof of Theorem 1.1.
4 Proof of Theorem 1.2
In this section, we discuss the case that f(0) is unbounded, that is, f has a singularity at \(u=0\) and can be defined as \(f=\frac{a(t)}{u^{\nu }}\) where \(\nu \in (0,\infty )\) is a constant. Then, since a is sign-changing, the maximum principle does not hold. Hence, we need to choose appropriate domain so that \(\lambda \frac{a(t)}{u^{\mu }}+u\) becomes positive. (1.1)-(1.2) can rewrite
where \(k'_1,k'_2,k'_3\) are constants and satisfy Lemma 2.3. Let
Now, we define two open sets
where \(r_1\) and \(r_2\) are two constants and
First, we show that \(\Phi (P\cap (\bar{\Omega }_2\backslash \Omega _1))\subset P\). In fact, for any \(u\in P\cap (\bar{\Omega }_2\backslash \Omega _1)\),
Since \(r_1>\frac{1}{\sigma }(\lambda \Vert a^{-}\Vert _{\infty })^{\frac{1}{\nu +1}}\), it follows that
By (4.2), then
which implies \(\Phi (P\cap (\bar{\Omega }_2\backslash \Omega _1))\subset P\). By using Arzela–Ascoli theorem, it is easy to show that \(\Phi \) is a completely continuous operator.
Next, we show that
In fact, for all \(u\in P\cap \partial \Omega _2\), we obtain that \(\Vert u\Vert _{\infty }=r_2\) and
From (4.2), then
Then we can choose that \(r_2\) larger enough such that
Therefore, \(\Vert \Phi u\Vert _{\infty }\le \Vert u\Vert _{\infty }\).
Finally, we show that
In fact, for all \(u\in P\cap \partial \Omega _1\), then \(\Vert u\Vert _{\infty }=r_1\) and
It follows from (4.2) that
since \(r_1=(\lambda \omega \sigma _*\Vert a^{+}\Vert _{\infty })^{\frac{1}{\nu +1}}\). Hence, \(\Vert \Phi u\Vert _{\infty }\ge \Vert u\Vert _{\infty }\) holds.
In conclusion, (1.1)–(1.2) has an \(\omega \)-periodic solution u satisfying \(u\in [\sigma r_1,r_2]\). Furthermore,
so there exists a constant \(\lambda _0\) such that (1.1)–(1.2) has at least one positive periodic solution for all \(\lambda \in (0,\lambda _0)\).
References
Kong, L.B., Wang, S.T., Wang, J.Y.: Positive solution of a singular nonlinear third-order periodic boundary value problems. J. Comput. Appl. Math. 132, 247–253 (2001)
Sun, J., Liu, Y.: Multiple positive solutions of singular third-order periodic boundary value problem. Acta Math. Sci. Ser. B Engl. Ed. 25, 81–88 (2005)
Chu, J.F., Zhou, Z.C.: Positive solutions for singular non-linear third-order periodic boundary value problems. Nonlinear Anal. 64, 1528–1542 (2006)
Feng, Y.Q.: On the existence and multiplicity of positive periodic solutions of a nonlinear third-order equation. Appl. Math. Lett. 22, 1220–1224 (2009)
Chen, Y.L., Ren, J.L., Siegmund, S.: Green’s function for third-order differential equations. Rocky Mountain J. Math. 41, 1417–1448 (2011)
Ren, J.L., Cheng, Z.B., Chen, Y.L.: Existence results of periodic solutions for third-order nonlinear singular differential equation. Math. Nachr. 286, 1022–1042 (2013)
Hess, P., Kato, T.: On some linear and nonlinear eigenvalue problems with an indefinite weight function. Comm. Part. Differ. Equ. 5, 999–1030 (1980)
Hai, D.D.: Positive solutions to a class of elliptic boundary value problems. J. Math. Anal. Appl. 227, 195–199 (1998)
Bravo, J.L., Torres, P.J.: Periodic solutions of a singular equation with indefinite weight. Adv. Nonlinear Stud. 10, 927–938 (2010)
Cheng, Z.B., Cui, X.X.: Positive periodic solution to an indefinite singular equation. Appl. Math. Lett. 112, 7 (2021)
Han, X.F., Cheng, Z.B.: Positive periodic solutions to a second-order singular differential equation with indefinite weights. Qual. Theory Dyn. Syst. 21, 16 (2022)
Hakl, R., Zamora, M.: Periodic solutions to second-order indefinite singular equations. J. Differ. Equ. 263, 451–469 (2017)
Godoy, J., Zamora, M.: A general result to the existence of a periodic solution to an indefinite equation with a weak singularity. J. Dyn. Differ. Equ. 31, 451–468 (2019)
Cheung, W.S., Ren, J.L., Han, W.W.: Positive periodic solution of second order neutral functional differential equations. Nonlinear Anal. 71, 3948–3955 (2009)
Ma, R.Y., An, Y.L.: Global strcture of positive solutions for superlinear second-order \(m\)-point boundary value problems. Topol. Methods Nonlinear Anal. 3, 279–290 (2009)
Deimling, K.: Nonlinear Functional Analysis. Spring, New York (1985)
Dancer, E.: Global solutions branches for positive mapsm Arch. Ration. Mech. Anal. 55, 207–213 (1974)
Zeidler, E.: Nonlinear Functional Analysis and Its Applications. Springer, New York (1986)
Elias, U.: Eigenvalue problems for the equations \(Ly+\lambda p(x)y=0\). J. Differ. Equ. 29, 28–57 (1978)
Acknowledgements
The authors are grateful to the anonymous referees for their constructive comments and suggestions which have greatly improved this paper.
Funding
This work was supported by the National Natural Science Foundation of China (Grant Nos. 61877046).
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that there is no conflict of interests regarding the publication of this article.
Additional information
Communicated by Pei Yu.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Liu, S., Bi, Z. Singular and Nonsingular Third-Order Periodic Boundary Value Problems with Indefinite Weight. Bull. Iran. Math. Soc. 49, 44 (2023). https://doi.org/10.1007/s41980-023-00789-1
Received:
Revised:
Accepted:
Published:
DOI: https://doi.org/10.1007/s41980-023-00789-1
Keywords
- Positive solutions
- Indefinite weight
- Singularity
- Third-order periodic boundary value problem
- Bifurcation