数学物理学报  2018, Vol. 38 Issue (4): 770-778   PDF    
扩展功能
加入收藏夹
复制引文信息
加入引用管理器
Email Alert
RSS
本文作者相关文章
沈文国
一类p-Laplacian方程单侧全局区间分歧及应用
沈文国     
兰州工业学院基础学科部 兰州 730050
摘要:首先建立一类含不可微非线性项p-Laplacian方程的单侧全局区间分歧定理.应用上述定理,可以证明一类半线性p-Laplacian方程主半特征值的存在性.进而,可研究下列半线性p-Laplacian方程结点解的存在性 $ \begin{eqnarray} \left\{ \begin{array}{ll} -(r^{N-1}\varphi_{p}(u'))'=\alpha(r)r^{N-1}\varphi_{p}(u^{+})+\beta(r)r^{N-1} \varphi_{p}(u^{-})\\ \textrm{}\ \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, +\lambda a(r)r^{N-1}f(u), \, \, {\rm {\rm a.e.}}\, \, r\in(0, 1),\\ \textrm{}\ u'(0)=u(1)=0, \end{array} \right.\nonumber \end{eqnarray} $ 其中$1 < p < +\infty$, $\varphi_{p}(s)=|s|^{p-2}s$, $a(r)\in C[0, 1]$, $a(r)\geq0$且在$[0, 1]$的任何子集上成立$a(r)\not\equiv0$; $\lambda$是一个参数, $u^{+} = \max\{u, 0\}, u^{-}= -\min\{u, 0\}$, $\alpha, \beta\in C[0, 1]$; 对于$s\in{\Bbb R}^{+}, $都有$f\in C({\Bbb R}, {\Bbb R})$$sf(s)>0$, ${\Bbb R}^+=[0, +\infty)$, 并且满足$f_{0}\in [0, \infty)$$f_{\infty}\in (0, \infty)$或者$f_{0}\in (0, \infty]$$f_{\infty}=0$或者$f_{0}=0$$f_{\infty}=\infty, $其中$f_{0}= \lim\limits_{|s| \rightarrow0} f(s)/s, f_{\infty}=\lim\limits_{|s| \rightarrow+\infty} f(s)/s.$该文用单侧全局分歧技巧和连通分支极限证明结论.
关键词单侧全局区间分歧    半线性问题    结点解    p-Laplacian方程    
Interval Bifurcation for the p-Laplacian Equation and Its Applications
Shen Wenguo     
Department of Basic Courses, Lanzhou Institute of Technology, Lanzhou 730050
Abstract: In this paper, we establish a unilateral global interval bifurcation result for the p-Laplacian equation. Furthermore, we shall prove the existence of the principal half-eigenvalues for the half-linear p-Laplacian equation. Moreover, we also investigate the existence of radial nodal solutions for the problems. $ \begin{eqnarray} \left\{ \begin{array}{ll} -(r^{N-1}\varphi_{p}(u'))'=\alpha(r)r^{N-1}\varphi_{p}(u^{+})+\beta(r)r^{N-1} \varphi_{p}(u^{-})\\ \textrm{}\ \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, +\lambda a(r)r^{N-1}f(u), \, \, {\rm {\rm a.e.}}\, \, r\in(0, 1),\\ \textrm{}\ u'(0)=u(1)=0, \end{array} \right.\nonumber \end{eqnarray} $ where $1 < p < +\infty$, $\varphi_{p}(s)=|s|^{p-2}s$, $a(r)\in C[0, 1]$, $a(r)\geq0, $ $a(r)\not\equiv0$ on any subinterval of $[0, 1]$; $\lambda$ is a parameter, $u^{+} = \max\{u, 0\}, u^{-}= -\min\{u, 0\}$, $\alpha, \beta\in C[0, 1]$ are radially symmetric; $f\in C({\Bbb R}, {\Bbb R})$, $sf(s)>0$ for $s\in{\Bbb R}^{+}, $ and $f_{0}\in [0, \infty)$ and $f_{\infty}\in (0, \infty)$ or $f_{0}\in (0, \infty]$ and $f_{\infty}=0$ or $f_{0}=0$ and $f_{\infty}=\infty, $ where $f_{0}= \lim\limits_{|s| \rightarrow0} f(s)/s, f_{\infty}=\lim\limits_{|s| \rightarrow+\infty} f(s)/s.$ We use unilateral global bifurcation techniques and the approximation of connected components to prove our main results.
Key words: Unilateral interval bifurcation     Half-quasilinear problems     Nodal solutions     p-Laplacian equation    
1 引言

在文献[1-3]中, 作者研究了非线性问题分歧理论.近来, 应用文献[4], 代国伟等[5-6]亦研究了$p$-Laplacian方程. 1977年, Berestycki[7]研究了一类含不可微非线性项的二阶问题.最近, 代国伟等[8-9]也考虑了相似于文献[7]的问题, 作者[10]亦考虑了相似的周期问题.受上述文献启发, 本文将研究下列问题

$ \begin{eqnarray} \left\{ \begin{array}{ll} -{\rm div}(\varphi_{p}(\nabla u))=\lambda a(x)\varphi_{p}(u)+F(x, u, \lambda), ~~ &{\rm in}\, B, \\ u(x)=0, &{\rm on}\, \partial B, \end{array} \right. \end{eqnarray} $ (1.1)

其中$B$${\mathbb{R}}^{N}$上的单位球, $F=f+g$, $f, g\in C(\overline{B}\times {\mathbb{R}}^{2})$$a(x)\in C(\overline{B})$关于$x$是径向对称的.

显然, 问题(1.1)的径向解等价于下列问题的解

$ \begin{eqnarray} \left\{ \begin{array}{ll} -(r^{N-1}\varphi_{p}(u'))'=\lambda r^{N-1} a(r)\varphi_{p}(u)+r^{N-1}F(r, u, \lambda), \, \, {\rm {\rm a.e.}}\, \, r\in(0, 1),&\textrm{}\\ u'(0)=u(1)=0, \end{array} \right. \end{eqnarray} $ (1.2)

其中$F=f+g$, $f, g\in C([0, 1]\times {\mathbb{R}}^{2})$, $a\in C[0, 1]$, 满足

(H0) $a(x)\in C(\overline{B})$是径向对称的, 对于$x\in\overline{B}$, $r=|x|$, 都有$a(r)\geq0, $且在$[0, 1]$的任何子集上满足$a(r)\not\equiv0$;

(H1)对任何$r\in(0, 1)$, $\lambda\in {\mathbb{R}}$, $0<\vert s\vert\leq1$, 都有$\left|\frac{f(r, s, \lambda)}{\varphi_{p}(s)}\right|\leq M$, 其中$M$是一个正常数;

(H2)对于$s=0$, $r\in(0, 1)$, $\lambda$在有界集中, 都有$g(r, s, \lambda)=o(\vert \varphi_{p}(s)\vert)$.

另外, 半线性或半拟线性问题[7-10]已经引起人们的关注.在上述单侧全局区间分歧结果的基础上, 接着研究如下半线性特征值问题

$ \begin{eqnarray} \left\{ \begin{array}{ll} -{\rm div}(\varphi_{p}(\nabla u))=\lambda a(x)\varphi_{p}(u)+\alpha(x)\varphi_{p} (u^{+})+\beta(x) \varphi_{p}(u^{-}), ~~&{\rm in}\, B, \\ u(x)=0, &{\rm on}\, \partial B, \end{array} \right. \end{eqnarray} $ (1.3)

其中$\alpha(x), \beta(x)\in C(\overline{B})$是径向对称的.

显然, 问题(1.3)的径向解等价于下列问题的解

$ \begin{eqnarray} \left\{ \begin{array}{ll} -(r^{N-1}\varphi_{p}(u'))'= r^{N-1}(\lambda a(r)\varphi_{p}(u)+\alpha(r)\varphi_{p}(u^{+})+\beta(r)\varphi_{p}(u^{-})), \, \, {\rm {\rm a.e.}}\, \, r\in(0, 1),&\textrm{}\\ u'(0)=u(1)=0, \end{array} \right. \end{eqnarray} $ (1.4)

其中$u^{+} = \max\{u, 0\}, u^{-}= -\min\{u, 0\}$, $\alpha(r), \beta(r)$满足

(H3) $\alpha(r), \beta(r)\in C[0, 1]$是径向对称的.

在上述特征值理论的基础上,接着研究下列半线性问题径向结点解的存在性问题

$ \begin{eqnarray} \left\{ \begin{array}{ll} -{\rm div}(\varphi_{p}(\nabla u))=\alpha(x)\varphi_{p}(u^{+})+\beta(x) \varphi_{p}(u^{-})+\lambda a(x)f(u)), \, \,&{\rm in}\, B, \\ u(x)=0, &{\rm on}\, \partial B. \end{array} \right. \end{eqnarray} $ (1.5)

显然, 问题(1.5)的径向解等价于下列问题的解

$ \begin{eqnarray} \left\{ \begin{array}{ll} -(r^{N-1}\varphi_{p}(u'))'=r^{N-1}(\alpha(r)\varphi_{p}(u^{+})+\beta(r)\varphi_{p}(u^{-})+\lambda a(r)f(u)), \, \, {\rm {\rm a.e.}}\, \, r\in(0, 1),&\textrm{}\\ u'(0)=u(1)=0. \end{array} \right. \end{eqnarray} $ (1.6)

假设$f$满足

(H4) $sf(s)>0$, $s\neq0;$

(H5) $f_{0}, f_{\infty}\in (0, +\infty);$

(H6) $f_{0}=0$, $f_{\infty}\in (0, \infty);$

(H7) $f_{0}\in (0, \infty)$, $f_{\infty}=0;$

(H8) $f_{0}=0$, $f_{\infty}=\infty;$

(H9) $f_{0}=\infty$, $f_{\infty}=0.$

其中

$ f_{0}= \lim\limits_{|s| \rightarrow0}\frac{f(s)}{\varphi_{p}(s)}, \;\;\; f_{\infty}=\lim\limits_{|s| \rightarrow+\infty}\frac{f(s)}{\varphi_{p}(s)}. $
2 预备知识

$Y=C[0, 1]$伴随着范数$\|u\|_{\infty}=\max\limits_{r\in[0, 1]}|u(r)|$.设$E:=\{u(r) \in C^{1}[0, 1]|u'(0)=u(1)=0\}$伴随着通常范数$\|u\|=\max\limits_{r\in [0, 1]}|u(r)|+\max\limits_{r\in [0, 1]}|u'(r)|$.用$S_{k}^{+}$表示$E$中恰有$k-1$个内部非退化零点且在$r=0$附近取正值的函数集合, 又设$S_{k}^{-}=-S_{k}^{+}$$S_{k}=S_{k}^{+}\cup S_{k}^{-}.$显然$S_{k}^{+}$$S_{k}^{-}$$E$中不相交开集.记$\Phi_{k}^{\pm}={\mathbb{R}}\times S_{k}^{\pm}, $ $\Phi_{k}={\mathbb{R}}\times S_{k}$.最后, 用${\cal S}$表示问题(1.2)在${\mathbb{R}}\times E$中非平凡解集的闭包, 并且当$u\in \Phi_{k}^{\pm}$时, 用${\cal S}_{k}^{\pm}$表示${\cal S}$的子集且${\cal S}_{k}={\cal S}_{k}^{+}\cup {\cal S}_{k}^{-}.$

$E$是一个实Banach空间且装备范数$\|\cdot\|$.考虑算子方程

$ \begin{equation} u=\lambda Bu+H(\lambda, u), \end{equation} $ (2.1)

其中$B$是一个紧线性算子, $H:{\mathbb{R}}\times E\rightarrow E$是紧的且在$u=0$$\lambda$在有界区间中, $H=o(\|u\|)$一致成立.假如$B$的特征值$\mu$重数为$1$${\cal S}$表示问题(2.1)的非平凡解集.由Dancer[4]可知, 存在两个不同的无界连续统${\cal C}^{+}_{\mu}$${\cal C}^{-}_{\mu}$组成了从$(\mu, 0)$发出的${\cal S}$的分歧${\cal C}_{\mu}$, 且${\cal C}^{+}_{\mu}\cap {\cal C}^{-}_{\mu}\neq\{(\mu, 0)\}$.

本节主要定理如下

定理 2.1  假设(H0), (H1), (H2)成立.令$d=M/a_{0}$, 其中$a_{0}=\min\limits_{r\in[0, 1]}a(r)$, 且令$I_{k}=[\lambda_{k}-d, \lambda_{k}+d]$, 则${\cal S}_{k}^{\nu}\cup(I_{k}\times\{0\})$的包含$I_{k}\times \{0\}$的连通分支${\cal C}_{k}^{\nu}$是无界的且属于$({\mathbb{R}}\times S_{k}^{\nu})\cup(I_{k}\times\{0\})$, 其中$\nu=+, -$.

为了证明定理2.1, 考虑下列辅助问题

$ \begin{eqnarray} \left\{ \begin{array}{ll} (r^{N-1}\varphi_{p}(u'))'+r^{N-1}(\gamma a(r)\varphi_{p}(u)+f(r, u|u|^{\epsilon}, \lambda)+g(r, u, \lambda))=0, \, \, {\rm a.e.}\, \, r\in(0, 1),&\textrm{}\\ u'(0)=u(1)=0.& \textrm{} \end{array} \right. \end{eqnarray} $ (2.2)

由文献[11, 382], 可得下列结果

引理 2.1  令$L[y]=(r^{N-1}\varphi_{p}(y'))'$, $y(\xi)=y(\eta)=0$, $y(t)>0, z(t)>0, t\in(\xi, \eta)$, 可得

$ \int_{\xi}^{\eta}\frac{y}{\varphi_{p}(z)}\left(\varphi_{p}(z)L[y]-\varphi_{p}(y)L[z]\right){\rm d}r\leq0. $

先证如下关键引理成立.

引理 2.2  令$\epsilon_{n}\rightarrow 0$, $0<\epsilon_{n}<1$.假如存在序列$(\lambda_{n}, u_{n})\in {\mathbb{R}}\times S_{k}^{\nu}$使得$(\lambda_{n}, u_{n})$是问题(2.2)对应于$\epsilon=\epsilon_{n}$的一个非平凡解, 且在${\mathbb{R}}\times E$$(\lambda_{n}, x_{n})\rightarrow (\lambda, 0)$, 则$\lambda\in I_{k}$.

  不失一般性, 假设$\|u_{n}\|\leq 1$.令$w_{n}=u_{n}/\|u_{n}\|$, 则$w_{n}$满足

$ \begin{eqnarray} \left\{ \begin{array}{ll} (r^{N-1}\varphi_{p}(w_{n}'))'+\lambda_{n} r^{N-1}a(r)\varphi_{p}(w_{n})+\frac{r^{N-1}f(r, u_{n}|u_{n}|^{\epsilon_{n}}, \lambda_{n})}{\|u_{n}\|^{p-1}}& \textrm{}\\[3mm] +\frac{r^{N-1}g(r, u_{n}, \lambda_{n})}{\|u_{n}\|^{p-1}}=0, \, \, {\rm a.e.}\, \, r\in(0, 1),&\textrm{}\\ w_{n}'(0)=w_{n}(1)=0.& \textrm{}\nonumber \end{array} \right. \end{eqnarray} $

相似于文献[9]中引理2.3的证明, 当$\|u\|\rightarrow0$, $r\in (0, 1)$, 且$\lambda$在有界集中时, 可得$|g(r, u, \lambda)|/\|u\|^{p-1}\rightarrow0$; 并且对所有$r\in (0, 1)$, 可得$|f(r, u_{n}|u_{n}|^{\epsilon_{n}}, \lambda_{n})|/\|u_{n}\|^{p-1}\rightarrow M$.接着, 在$E$中,可得$w_{n}\rightarrow w$$\|w\|=1$.显然, 可得$w\in \overline{S_{k}^{\nu}}$.可以断言$w\in S_{k}^{\nu}$.反设$w\in \partial S_{k}^{\nu}$, 由文献[5]中引理2.2, 可知$w\equiv0$, 与$\|w\|=1$矛盾.

由于$w_{n}, \psi_{k}^{\nu}\in S_{k}^{\nu}$, 由文献[7]中引理2可知, 存在(0, 1)中两个子区间$[\xi_{1}, \eta_{1}]$$[\xi_{2}, \eta_{2}]$, 使得$w_{n}$$\psi_{k}^{\nu}$$[\xi_{1}, \eta_{1}]$$[\xi_{2}, \eta_{2}]$中有定义并且具有相同的符号, 同时下满足$w_{n}(\xi_{1})=w_{n}(\eta_{1})=0$, $\psi_{k}^{\nu}(\xi_{2})=\psi_{k}^{\nu}(\eta_{2})=0$.

不失一般性, 假设在$(\xi_{1}, \eta_{1})$中, $w_{n}>0$$\psi_{k}^{\nu}>0$.由引理2.1可得

$ \begin{eqnarray} &&\int_{\xi_{1}}^{\eta_{1}}\frac{w_{n}}{\varphi_{p}(\psi_{k}^{\nu})}\left(\varphi_{p}(\psi_{k}^{\nu})L[w_{n}]-\varphi_{p}(w_{n})L[\psi_{k}^{\nu}]\right){\rm d}r \\&=&\int_{\xi_{1}}^{\eta_{1}}\left[(\lambda_{k}-\lambda_{n})a(r) -\frac{f(r, u_{n}|u_{n}|^{\epsilon_{n}}, \lambda_{n})}{\|u_{n}\|^{p-1}\varphi_{p}(w_{n})}-\frac{g(r, u_{n}, \lambda_{n})}{\|u_{n}\|^{p-1}\varphi_{p}(w_{n})}\right]r^{N-1}w_{n}^{p}{\rm d}r\leq0. \end{eqnarray} $ (2.3)

相似的, 可得

$ \begin{equation} \int_{\xi_{2}}^{\eta_{2}}\left[(\lambda_{n}-\lambda_{k})a(r) -\frac{f(r, u_{n}|u_{n}|^{\epsilon_{n}}, \lambda_{n})}{\|u_{n}\|^{p-1}\varphi_{p}(w_{n})}-\frac{g(r, u_{n}, \lambda_{n})}{\|u_{n}\|^{p-1}\varphi_{p}(w_{n})}\right]r^{N-1}\psi_{k}^{\nu}{\rm d}r\leq0. \end{equation} $ (2.4)

假如$\lambda\leq\lambda_{k}$, 由(2.3)式, (H1)和(H2), 可得

$ \begin{eqnarray*} \int_{\xi_{1}}^{\eta_{1}}(\lambda_{k}-\lambda)a(r) r^{N-1}w^{p} {\rm d}r& \leq& \lim\limits_{n \rightarrow\infty}\int_{\xi_{1}}^{\eta_{1}}\frac{f(r, u_{n}|u_{n}|^{\epsilon_{n}}, \lambda_{n})}{\|u_{n}\|^{p-1}}r^{N-1}w_{n}^{p}{\rm d}r \\&\leq&\lim\limits_{n\rightarrow\infty}\int_{\xi_{1}}^{\eta_{1}}\frac{f(r, u_{n}|u_{n}|^{\epsilon_{n}}, \lambda_{n})}{\varphi_{p}(u_{n}|u_{n}|^{\epsilon_{n}})}\varphi_{p}(|u_{n}|^{\epsilon_{n}})r^{N-1}w_{n}^{p}{\rm d}r \\&\leq&\lim\limits_{n\rightarrow\infty}\int_{\xi_{1}}^{\eta_{1}}\frac{f(r, u_{n}|u_{n}|^{\epsilon_{n}}, \lambda_{n})}{\varphi_{p}(u_{n}|u_{n}|^{\epsilon_{n}})}\varphi_{p}(\frac{|u_{n}|^{\epsilon_{n}}}{\|u_{n}\|^{\epsilon_{n}}})\varphi_{p}(\|u_{n}\|^{\epsilon_{n}})r^{N-1}w_{n}^{p}{\rm d}r \\&\leq&\int_{\xi_{1}}^{\eta_{1}}Mr^{N-1}w^{p}{\rm d}r.\nonumber \end{eqnarray*} $

因此, 可知

$ \int_{\xi_{1}}^{\eta_{1}}(\lambda_{k}-\lambda)a_{0} r^{N-1}w^{p} {\rm d}r \leq\int_{\xi_{1}}^{\eta_{1}}Mr^{N-1}w^{p}{\rm d}r, \nonumber $

进而, 可得$\lambda\geq\lambda_{k}-d$.

假设$\lambda\geq\lambda_{k}$, 由(2.4)式, (H1)和(H2), 可得

$ \begin{eqnarray*} \int_{\xi_{2}}^{\eta_{2}}(\lambda-\lambda_{k})a(r) r^{N-1}\psi_{k}^{\nu} {\rm d}r&\leq&\lim\limits_{n \rightarrow\infty}\int_{\xi_{2}}^{\eta_{2}}\frac{-f(r, u_{n}|u_{n}|^{\epsilon_{n}}, \lambda_{n})}{\|u_{n}\|^{p-1}\varphi_{p}(w_{n})}r^{N-1}\psi_{k}^{\nu}{\rm d}r \\&\leq&\int_{\xi_{2}}^{\eta_{2}}Mr^{N-1}\psi_{k}^{\nu}{\rm d}r.\nonumber \end{eqnarray*} $

因此, 可知

$ \int_{\xi_{2}}^{\eta_{2}}(\lambda-\lambda_{k})a_{0} r^{N-1}\psi_{k}^{\nu} {\rm d}r\leq\int_{\xi_{2}}^{\eta_{2}}Mr^{N-1}\psi_{k}^{\nu}{\rm d}r, \nonumber $

进而, 可得$\lambda\leq\lambda_{k}-d$.

因此, 可得$\lambda\in I_{k}$.引理2.2证毕.

定理2.1的证明  仅证${\cal C}_{k}^{+}$的情况, 同理可证${\cal C}_{k}^{-}$.令${\cal C}_{k}^{+}$${\cal S}_{k}^{+}\cup(I_{k}\times\{0\})$的包含$I_{k}\times \{0\}$的紧连通分支.

应用与文献[9]中定理1.1相似的证明, 可得${\cal C}_{k}^{+}$无界且${\cal C}_{k}^{+}\subset({\mathbb{R}}\times S_{k}^{+})\cup(I_{k}\times\{0\})$.

考虑问题$(2.1)$.对于$\epsilon>0$, 易知$f(r, u|u|^{\epsilon}, \lambda)+g(r, u, \lambda)$满足条件(H2).令${\cal S}_{\epsilon}$是问题(2.2)的非平凡解.由文献[6]中定理3.1-3.2, 存在两个从$(\lambda_{k}^{\nu}, 0)$分歧出的无界连通分支${\cal C}_{k, \epsilon}^{+}\subseteq{\cal S}_{\epsilon}^{+}$${\cal C}_{k, \epsilon}^{-}\subseteq{\cal S}_{\epsilon}^{-}$, 使得${\cal C}_{k, \epsilon}^{\nu}\subset({\mathbb{R}}\times S_{k}^{\nu})\cup \{(\lambda_{k}^{\nu}, 0)\}, $$\nu=+$$-.$因此, 对所有$\epsilon>0$存在$(\lambda_{\epsilon}, u_{\epsilon})\in {\cal C}^{+}_{k, \epsilon}\cap\partial{\cal O}$.因为在${\mathbb{R}} \times S_{k}^{+}$${\cal O}$是无界的, 由方程(2.2)可知$(\lambda_{\epsilon}, u_{\epsilon})$${\mathbb{R}}\times C^{2}$中是有界且不依赖与$\epsilon$.由于$G_{p}$是紧的, 可找到序列$\epsilon_{n}\rightarrow0$使得$(\lambda_{\epsilon_{n}}, u_{\epsilon_{n}})$收敛于(1.2)式的解$(\lambda, u)$.因此$u\in \overline{S_{k}^{+}}$.假如$u\in \partial S^{+}_{k}$, 则由文献[4]中引理2.2可得$u\equiv0$.由引理2.2, $\lambda\in I_{k}, $${\cal O}$的定义矛盾.另外, 假如$u\in S_{k}^{+}$, 则$(\lambda, u)\in {\cal S}_{k}^{+}\cap\partial{\cal O}$, 与${\cal S}_{k}^{+}\cap\partial{\cal O}=\emptyset$相矛盾.证毕.

由定理2.1及证明, 易得

推论 2.1  在${\mathbb{R}}\times E$中, 问题(1.2)存在两个从$I_{k}\times\{0\}$分歧出的解的无界连通分支${\cal D}_{k}^{+}$${\cal D}_{k}^{-}$, 使得${\cal D}_{k}^{\nu}\subset({\mathbb{R}}\times S_{k}^{\nu})\cup(I_{k}\times\{0\})$, 其中$\nu=+, -$

接着, 将考虑半拟线性问题(1.4).由于在锥$u > 0$$u<0$中, 问题(1.4)是正的$p-1$齐次的且是线性的, 故称之为半拟线性的.相似于文献[6], 假若问题(1.4)存在非平凡解$(\lambda, u_{\lambda})$, 则称$\lambda$是问题(1.4)的半特征值.假若对问题(1.4)的所有解$(\lambda, v)$都有$v=cu_{\lambda}, c > 0$, 则称$\lambda$是简单的.假若半特征值对应的特征函数是正的或负的, 则称之为主半特征值.

为了证明定理3.1, 需要得出下列比较结果.

引理 2.3  对于$r\in(0, 1)$$b_{i}(r)\in C(0, 1), i = 1, 2$, 令$b_{2}(r)\geq \max\{b_{1}(r), b_{1}(r)+\alpha(r)+\beta(r), b_{1}(r)-\alpha(r)-\beta(r)\}$.令$u_{1}, u_{2}$分别是下列方程的解

$ -(r^{N-1}\varphi_{p}(u'))'=b_{i}(r)r^{N-1}\varphi_{p}(u)+\alpha(r)r^{N-1} \varphi_{p}(u^{+})+\beta(r)r^{N-1} \varphi_{p}(u^{-}) , r\in(0, 1), i = 1, 2, $

假如$(c, d)\subseteq(0, 1)$, $u_{1}(c) = u_{1}(d) = 0$, 对于$r\in (c, d)$, 满足$u_{1}(r)\neq0$, 则要么存在$\tau\in(c, d)$使得$u_{2}(\tau ) = 0$成立, 要么$b_{2} = b_{1}$$b_{2} = b_{1}+\alpha+\beta$$b_{2} = b_{1}-\alpha-\beta$并且对于常数$\mu\neq0$, $u_{2}(r) = \mu u_{1}(r)$成立.

  相似于文献[9]中定理3.1的证明方法, 可得结果.

由引理2.3, 可得

引理 2.4  令$I=[a, b]$使得$I\subset [0, 1]$, 且meas$(I) > 0.$$g_{n}\in C([0, 1], {\mathbb{R}})$使得

$ \lim\limits_{n\rightarrow+\infty}g_{n}(r)=+\infty, \, \, \forall r\in I. $

$y_{n}\in E$是下列方程的一个解

$ -(r^{N-1}\varphi_{p}(y_{n}'))'=b_{i}(r)r^{N-1}\varphi_{p}(y_{n})+\alpha(r)r^{N-1} \varphi_{p}(y_{n}^{+})+\beta(r)r^{N-1} \varphi_{p}(y_{n}^{-}) , r\in(0, 1), i = 1, 2. $

$n\rightarrow+\infty$时, 则$y_{n}$$I$上零点个数趋于无穷大.

  应用相似于文献[5]中引理3.2或文献[6]中引理2.2的证明方法, 结合引理2.3, 可得结果.

由定理2.1, 可建立问题(1.4)的谱如下

定理 2.2  问题(1.4)存在两个半线性特征值序列$\lambda_{1}^{+}<\lambda_{2}^{+}\cdots <\lambda_{k}^{+}<\cdots $$\lambda_{1}^{-}<\lambda_{2}^{-}\cdots <\lambda_{k}^{-}< \cdots $, 对应的半线性解分别在$\{\lambda_{k}^{+}\}\times S_{k}^{+}$$\{\lambda_{k}^{-}\}\times S_{k}^{-}$中, 且除了这些解和平凡解, 问题(1.4)没有别的半线性特征值解.

  应用相似于文献[9]中定理1.2的证明方法, 可得结果.

现考虑问题

$ \begin{eqnarray} \left\{ \begin{array}{ll} -(r^{N-1}\varphi_{p}(u'))'=\lambda a(r)r^{N-1}\varphi_{p}(u)+\alpha(r) r^{N-1} \varphi_{p}(u^{+})&\textrm{}\\ \;\;\, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, +\beta(r) r^{N-1}\varphi_{p}(u^{-})+r^{N-1}g(r, u, \lambda), \, \, {\rm a.e.}\, \, r\in(0, 1),&\textrm{}\\ u'(0)=u(1)=0,&\textrm{} \end{array} \right. \end{eqnarray} $ (2.5)

其中$g(r, u, \lambda)$满足(H2).可得如下结论

定理 2.3  对于$k\in{\Bbb N}$, $\nu=+, -$, $(\lambda_{k}^{\nu} , 0)$是问题(2.5)的一个分歧点.进而, 存在问题(2.5)的一个解的无界连通分支${\cal D}_{k}^{\nu}$, 使得${\cal D}_{k}^{\nu}\subset(({\mathbb{R}}\times S_{k}^{\nu})\cup\{(\lambda_{k}^{\nu}, 0)\})$.

  应用相似于文献[9]中定理3.3的证明方法, 可得结果.

3 半线性特征值问题的结点解

本节主要结论如下.

定理 3.1  假设(H0), (H3), (H4)和(H5)成立.对于$\nu=+, -$, 要么$ \frac{\lambda_{k}^{\nu}}{f_{\infty}}<\lambda<\frac{\lambda_{k}^{\nu}}{f_{0}}$成立, 要么$ \frac{\lambda_{k}^{\nu}}{f_{0}}<\lambda<\frac{\lambda_{k}^{\nu}}{f_{\infty}}$成立.则问题$(1.6)$存在两个解$u_{k}^{+}$, $u_{k}^{-}$, 使得$ u_{k}^{+}$在(0, 1)中存在$k-1$个简单零点且在$r=0$附近为正的;$ u_{k}^{-}$在(0, 1)中存在$k-1$个简单零点且在$r=0$附近为负的.

  应用文献[9]中, 相似于定理1.3的证明方法, 可得结果.

介绍下列引理

引理 3.1[12]  令$X$是一个Banach空间, 令$\{C_{n}|n=1, 2, \cdots \}$$X$的一个闭连通子序列集.假设

(ⅰ) 存在$z_{n}\in C_{n}, \, \, n=1, 2, \cdots $$z^{\ast}\in X$, 使得$z_{n}\rightarrow z^{\ast}$;

(ⅱ) $r_{n}=\sup\{\|x\||x\in C_{n}\}=\infty;$

(ⅲ) 对所有$R>0$, $\left(\bigcup\limits_{n=1}^{\infty}C_{n}\right)\cap B_{R}$$X$中的相对紧集, 其中

$ B_{R}=\{x\in X|\|x\|\leq R\}. $

则在${\mathbb{D}}$中,存在一个无界连通集$C$$z^{\ast}\in C, $其中${\mathbb{D}}:=\limsup\limits_{n\rightarrow\infty}C_{n}=\{x\in X|\exists\{n_{i}\}\subset{\Bbb N}$, $x_{n_{i}}\in C_{n_{i}}$, 使得$x_{n_{i}}\rightarrow x\}$ (可参考文献[35]).

定理 3.2  假设(H0), (H3), (H4)和(H6)成立.对于$\nu\in\{+, -\}$, 假设下列条件之一成立.

(ⅰ) 对于$\lambda_{k}^{\nu}>0$, 使得$\lambda \in(\frac{\lambda_{k}^{\nu}}{f_{\infty}}, +\infty)$成立;

(ⅱ) 对于$\nu\lambda_{k}^{\nu}>0, $使得$ \lambda \in(\frac{\lambda_{k}^{+}}{f_{\infty}}, +\infty)\cup(-\infty, \frac{\lambda_{k}^{-}}{f_{\infty}})$成立;

(ⅲ) 对于$\nu\lambda^{\nu}<0$, 使得$\lambda \in(\frac{\lambda_{k}^{-}}{f_{\infty}}, +\infty)\cup(-\infty, \frac{\lambda_{k}^{+}}{f_{\infty}})$成立;

(ⅳ) 对于$\lambda_{k}^{\nu}<0, $使得$\lambda \in(-\infty, \frac{\lambda_{k}^{\nu}}{f_{\infty}})$成立.

则问题$(1.6)$存在两个解$u_{k}^{+}$, $u_{k}^{-}$, 使得$ u_{k}^{+}$在(0, 1)中只有$k-1$个简单零点且在零点附近为正, $u_{k}^{-}$在(0, 1)中只有$k-1$个简单零点且在零点附近为负.

  仅证(ⅰ), 同理可证(ⅱ), (ⅲ)和(ⅳ).

由文献[13], 定义截断函数$f$如下

$ f^{[n]}(s):=\left\{ \begin{array}{ll} \frac{1}{n}\varphi_{p}(s), &s\in[-\frac{1}{n}, \frac{1}{n}], \\[3mm] \left[f(\frac{2}{n})-\frac{1}{n^{p}}\right](ns-2)+f(\frac{2}{n}),& s\in(\frac{1}{n}, \frac{2}{n}), \\[3mm] -\left[f(-\frac{2}{n})+\frac{1}{n^{p}}\right](ns+2)+f(-\frac{2}{n}), & s\in(-\frac{2}{n}, -\frac{1}{n}), \\[3mm] f(s), &s\in(-\infty, -\frac{2}{n}]\cup[\frac{2}{n}, +\infty). \end{array} \right. $

考虑

$ \begin{eqnarray} \left\{ \begin{array}{ll} -(r^{N-1}\varphi_{p}(u'))'=r^{N-1} \left(\alpha\varphi_{p}(u^{+})+\beta\varphi_{p}(u^{-})+\lambda a(r)f^{[n]}(u)\right) , \, \, {\rm {\rm a.e.}}\, \, r\in(0, 1),&\textrm{}\\ u'(0)=u(1)=0. \end{array} \right. \end{eqnarray} $ (3.1)

显然, 可得$\lim\limits_{n\rightarrow+\infty}f^{[n]}(s)= f(s)$, $(f^{[n]})_{0}=\frac{1}{n}$, $(f^{[n]})_{\infty}=f_{\infty}$.

由定理3.1可知, 存在从$(n\lambda_{k}^{\nu}, 0)$发出的问题(3.1)解的无界连续统${\cal D}_{k}^{\nu[n]}$, 并且连接到$(\frac{\lambda_{k}^{\nu}}{f_{\infty}}, \infty)$.

$z_{n}=(n\lambda_{k}^{\nu}, 0)$, $z^{\ast}=(\infty, 0), $$z_{n}\rightarrow z^{\ast}$.

因此, 引理3.1中(ⅰ)满足且$z^{\ast}=(\infty, 0)$.

显然, 引理3.1中(ⅱ)成立.由Arezela-Ascoli定理和$f^{[n]}$的定义可推得引理3.1中(ⅲ)成立.

因此, 由引理3.1可知, $\limsup\limits_{n\rightarrow\infty}{\cal D}_{k}^{\nu[n]}$ 包含一个无界连通分支${\cal D}_{k}^{\nu}$$(\infty, 0)\in {\cal D}_{k}^{\nu}$, $(\frac{\lambda_{k}^{\nu}}{f_{\infty}}, \infty)\in {\cal D}_{k}$.证毕.

定理 3.3  假设(H0), (H3), (H4)和(H7)成立.对于$\nu\in\{+, -\}$, 假设下列条件之一成立.

(ⅰ) 对于$\lambda_{k}^{\nu}>0$, 使得$\lambda\in(\frac{\lambda_{k}^{\nu}}{f_{0}}, +\infty)$成立;

(ⅱ) 对于$\nu\lambda_{k}^{\nu}>0$, 使得$\lambda\in(-\infty, \frac{\lambda_{k}^{-}}{f_{0}})\cup(\frac{\lambda_{k}^{+}}{f_{0}}, +\infty)$成立;

(ⅲ) 对于$\nu\lambda_{k}^{\nu}<0$, 使得$\lambda\in(-\infty, \frac{\lambda_{k}^{+}}{f_{0}})\cup(\frac{\lambda_{k}^{-}}{f_{0}}, +\infty)$成立;

(ⅳ) 对于$\lambda_{k}^{\nu}<0, $使得$\lambda\in(-\infty, \frac{\lambda_{k}^{\nu}}{f_{0}})$成立.

则问题$(1.6)$存在两个解$u_{k}^{+}$, $u_{k}^{-}$, 使得$ u_{k}^{+}$在(0, 1)中只有$k-1$个简单零点且在零点附近为正, $u_{k}^{-}$在(0, 1)中只有$k-1$个简单零点且在零点附近为负.

  仅证(ⅰ), 同理可证(ⅱ), (ⅲ)和(ⅳ).

假若$(\lambda, u)$是问题(1.6)的任意一个非平凡解, 在问题(1.6)两端同除以$\|u\|^{2(p-1)}$, 令$v =\frac{u}{\|u\|^{2}}$, 可得

$ \begin{eqnarray} \left\{ \begin{array}{ll} -(r^{N-1}\varphi_{p}(v'))'=r^{N-1} \left(\alpha\varphi_{p}(v^{+})+\beta\varphi_{p}(v^{-})+ \lambda a(r)\frac{f(u)}{\|u\|^{2(p-1)}}\right) , \, \, {\rm {\rm a.e.}}\, \, r\in(0, 1),&\textrm{}\\ u'(0)=u(1)=0. \end{array} \right. \end{eqnarray} $ (3.2)

定义

$ \widetilde{f}(v):=\left\{ \begin{array}{ll} \|v\|^{2(p-1)}f(\frac{v}{\|v\|^{2}}),&\mbox{若}~ v\neq0, \\ 0,&\mbox{若}~ v=0. \end{array} \right.\nonumber $

显然, 问题(3.2)等价于

$ \begin{eqnarray} \left\{ \begin{array}{ll} -(r^{N-1}\varphi_{p}(v'))'=r^{N-1} \left(\alpha\varphi_{p}(v^{+})+\beta\varphi_{p}(v^{-})+ \lambda a(r)\widetilde{f}(v)\right) , \, \, {\rm a.e.}\, \, r\in(0, 1),&\textrm{}\\ v'(0)=v(1)=0.& \textrm{} \end{array} \right. \end{eqnarray} $ (3.3)

显然$(\lambda, 0)$是问题$(3.3)$的一个解.简单计算可得$\widetilde{f}_{0} = f_{\infty}=0$, $\widetilde{f}_{\infty} = f_{0}\in(0, \infty)$.由定理3.2, 存在从$(\infty, 0)$发出的问题(3.3)解的一个无界连续统${\cal C}_{k}^{\nu}$, 使得${\cal C}_{k}^{\nu}\subset(({\mathbb{R}}\times S_{k}^{\nu})\cup\{(\infty, 0)\})$, 且${\cal C}_{k}^{\nu}$连接$(\infty, 0)$$(\frac{\lambda_{k}^{\nu}}{ \widetilde{f}_{\infty}}, \infty)$

由逆映射$v\rightarrow\frac{v}{\|v\|^{2}}=u$, 可得${\cal C}_{k}^{\nu}\rightarrow{\cal D}_{k}^{\nu}$是问题(1.6)解的一个无界连续统, 并且${\cal D}_{k}^{\nu}$连接$(\frac{\lambda_{k}^{\nu}}{{f}_{0}}, 0)$$(\infty, \infty)$.证毕.

定理 3.4  假设(H0), (H3), (H4)和(H8)成立.对于$\nu\in\{+, -\}$, 假设下列条件之一成立.

(ⅰ) 对于$\lambda_{k}^{\nu}>0$, 使得$\lambda\in(0, +\infty)$成立;

(ⅱ) 对于$\nu\lambda_{k}^{\nu}>0$$\nu\lambda_{k}^{\nu}<0$, 使得$\lambda\in(0, +\infty)\cup(-\infty, 0)$成立;

(ⅲ) 对于$\lambda_{k}^{\nu}<0$, 使得$\lambda\in(-\infty, 0)$成立.

则问题$(1.6)$存在两个解$u_{k}^{+}$, $u_{k}^{-}$, 使得$ u_{k}^{+}$在(0, 1)中只有$k-1$个简单零点且在零点附近为正, $u_{k}^{-}$在(0, 1)中只有$k-1$个简单零点且在零点附近为负.

  应用相似于定理3.2的证明方法, 可得问题(3.1).显然, 可得$\lim\limits_{n\rightarrow+\infty}f^{[n]}(s)= f(s)$, $(f^{[n]})_{0}=\frac{1}{n}$, $(f^{[n]})_{\infty}=f_{\infty}=\infty$.

由定理3.2, 存在从$(n\lambda_{k}^{\nu}, 0)$发出的问题(3.1)解的无界子连通分支${\cal D}_{k}^{\nu[n]}$, 且连接到$(0, \infty)$.

由引理3.1, 存在一个无界连续统${\cal D}$使得 $(\infty, 0)\in {\cal D}$ $(0, \infty)\in {\cal D}$.证毕.

定理 3.5  假设(H0), (H3), (H4)和(H9)成立.对于 $\nu\in\{+, -\}$ , 假设下列条件之一成立.

(ⅰ) 对于 $\lambda_{k}^{\nu}>0$ , 使得 $\lambda\in(0, +\infty)$ 成立;

(ⅱ) 对于 $\nu\lambda_{k}^{\nu}>0 或 \nu\lambda^{\nu}<0<$ , 使得 $\lambda\in(0, +\infty)\cup(-\infty, 0)$ 成立;

(ⅲ) 对于 $\lambda_{k}^{\nu}<0$ , 使得 $\lambda\in(-\infty, 0)$ 成立.

则问题(1.6)存在两个解 $u_{k}^{+}$ , $u_{k}^{-}$ , 使得 $u_{k}^{+}$ 在(0, 1)中只有 $k-1$ 个简单零点且在零点附近为正, $u_{k}^{-}$ 在(0, 1)中只有 $k-1$ 个简单零点且在零点附近为负.

  应用相似于定理3.3和定理3.4的证明方法, 可得结果.

注 3.1  由于 $\alpha u^{+}+\beta u^{-}$ , 问题(1.6)的非线性项在原点和无穷远处是线性的.显然, 文献[4-5]的分歧结果不能直接用来获得我们的结果.

注 3.2  当 $\alpha =\beta=0, f_{0}$ , $f_{\infty}\in (0, \infty)$ 时, Dai等[4-5]研究了问题(1.6), 可见, 定理5.1-5.5推广了文献[5]中定理3.1和文献[6]中定理4.1.

注 3.3  当 $p = 2$ , $N = 1$ , $a(x)>0$ , 且 $f_{0}, f_{\infty}\in (0, \infty).$ Ma等[8]和Dai等[9]分别研究了一维半线性问题的结点解.因此, 定理5.1-5.5推广了文献[8-9]中的结论.

参考文献
[1] Drábek P, Huang Y. Bifurcation problems for the p-Laplacian in ${\mathbb{R}}$N. Trans Amer Math Soc, 1997, 349: 171–188. DOI:10.1090/S0002-9947-97-01788-1
[2] Girg P, Takáč P. Bifurcations of positive and negative continua in quasilinear elliptic eigenvalue problems. Ann Inst H Poincaré Anal Non Linéaire, 2008, 9: 275–327.
[3] 马如云, 高承华. 二阶常微分方程周期解的全局分歧. 数学物理学报, 2009, 29A(5): 1223–1232.
Ma R Y, Gao C H. Global bifurcation of periodic solutions to second order ordinary differential equations. Acta Math Sci, 2009, 29A(5): 1223–1232.
[4] Dancer E N. On the structure of solutions of non-linear eigenvalue problems. Indiana Univ Math J, 1974, 23: 1069–1076. DOI:10.1512/iumj.1974.23.23087
[5] Dai G W, Ma R Y, Xu J. Global bifurcation and nodal solutions of N-dimensional p-Laplacian in unit ball. Appl Anal, 2013, 92(7): 1345–1356. DOI:10.1080/00036811.2012.678333
[6] Dai G W, Han X L, Ma R Y. Unilateral global bifurcation and nodal solutions for the p-Laplacian with sign-changing weight. Complex Var Elliptic Equ, 2014, 59(6): 847–862. DOI:10.1080/17476933.2013.791686
[7] Berestycki H. On some nonlinear Sturm-Liouville problems. J Differential Equations, 1977, 26: 375–390. DOI:10.1016/0022-0396(77)90086-9
[8] Ma R Y, Dai G W. Global bifurcation and nodal solutions for a Sturm-Liouville problem with a nonsmooth nonlinearity. J Funct Anal, 2013, 265: 1443–1459. DOI:10.1016/j.jfa.2013.06.017
[9] Dai G W, Ma R Y. Unilateral global bifurcation for p-Laplacian with non-p-1-linearization nonlinearity. Discrete Contin Dyn Syst, 2015, 35(1): 99–116.
[10] 沈文国. 一类半线性周期问题单侧全局区间分歧和定号解. 数学物理学报, 2017, 37A(5): 902–916.
Shen W G. Unilateral global interval bifurcation and one-sign solutions for the half-linear periodic problems. Acta Math Sci, 2017, 37A(5): 902–916. DOI:10.3969/j.issn.1003-3998.2017.05.011
[11] Kusano T, Jaros T, Yoshida N. A Picone-type identity and Sturmian comparison and oscillation theorems for a class of half-linear partial differential equations of second order. Nonlinear Anal, 2000, 40: 381–395. DOI:10.1016/S0362-546X(00)85023-3
[12] Ma R Y, An Y L. Global structure of positive solutions for nonlocal boundary value problems involving integral conditions. Nonlinear Anal, 2009, 71: 4364–4376. DOI:10.1016/j.na.2009.02.113
[13] Ambrosetti A, Calahorrano R M, Dobarro F R. Global branching for discontinuous problems. Comment Math Univ Carolin, 1990, 31: 213–222.