1 引言
该文研究了迁移率互异的可压电扩散模型PNPNS(Planck Nernest Poisson Navier Stokes) 的拟中性极限[1 -2 ] . 该系统为
(1.1) $\begin{equation} n_t^\lambda = {\rm div}(\mu_n (\nabla n^\lambda - n^\lambda \nabla \phi^\lambda ) - n^\lambda v^\lambda ), \label{eq:eqnp1} \end{equation}$
(1.2) $\begin{equation} p_t^\lambda = {\rm div}(\mu_p (\nabla p^\lambda + p^\lambda \nabla \phi^\lambda ) - p^\lambda v^\lambda ),\label{eq:eqnp2} \end{equation}$
(1.3) $\begin{equation} -\lambda^2 {\rm div}E^\lambda = n^\lambda - p^\lambda -D(x),\label{eq:eqnp3} \end{equation}$
(1.4) $ \begin{equation} (\rho^\lambda v^\lambda)_t+{\rm div}(\rho^\lambda v^\lambda \otimes v^\lambda)+A \nabla(\rho^\lambda)^2=\mu \Delta v^\lambda+\nabla{\rm div}v^\lambda+(n^\lambda-p^\lambda)\nabla \phi^\lambda, \label{eq:eqnp4} \end{equation}$
(1.5) $\begin{equation} \rho^\lambda_t+{\rm div}(\rho^\lambda v^\lambda)=0,\label{eq:eqnp5} \end{equation}$
(1.6) $\begin{equation} (n^\lambda,p^\lambda,v^\lambda,\rho^\lambda)(x,0)=(n_0^\lambda(x),p_0^\lambda(x),v_0^\lambda(x),\rho_0^\lambda(x)),\label{eq:eqnp7} \end{equation} $
其中$x\in {\Bbb T}^3, t>0$ , ${\Bbb T}^3$ 是${\Bbb R}^3$ 中的周期区域; 函数$n^\lambda$ , $p^\lambda$ , $\phi^\lambda$ , $v^\lambda$ , $\rho^\lambda$ 分别为负电荷浓度, 正电荷浓度, 电势, 电解液的速度和压力. 令$E^\lambda = - \nabla\phi^\lambda$ 为电场. 参数$\lambda>0$ 表示标量化的Debye长度, 通常很小. 参数$\mu$ $>$ $0$ 为粘性系数. $D(x)$ 为已知的掺杂分布函数. 参数$\mu_n$ , $\mu_p$ 为互异的迁移率, 且都为常数. $A$ 为任意的常数. 初始值$n_0^\lambda(x)$ , $p_0^\lambda(x)$ , $v_0^\lambda(x)$ 为光滑函数, 满足
(1.7) $\begin{equation} \int_{{\Bbb T}^3} \left(n_0^\lambda-p_0^\lambda-D(x)\right) {\rm d}x=0.\label{eq:eqnp8} \end{equation} $
(1.8) $\begin{equation} {\rm div}(\vec{a}\otimes\bar{b})=({\rm div}\vec{a})\otimes\vec{b}+(a\cdot\bigtriangledown)\vec{b}, \label{eq:fab1} \end{equation} $
(1.9) $\begin{equation} \rho^\lambda v^\lambda_t+\rho^\lambda v^\lambda\cdot \bigtriangledown v^\lambda+A \nabla(\rho^\lambda)^2=\mu \Delta v^\lambda+\nabla{\rm div}v^\lambda+(n^\lambda-p^\lambda)\nabla \phi^\lambda.\label{eq:eqnp4t1} \end{equation}$
假设, 当$\lambda\rightarrow 0$ 时, 有$(n^\lambda, p^\lambda, E^\lambda, \rho^\lambda,v^\lambda )$ $\rightarrow$ $(n, p, \varepsilon, \rho,v)$ 成立, 其中 $\varepsilon=-\nabla\phi$ , 那么形式上系统(1.1)-(1.3),(1.5),(1.9)式有极限系统为
(1.10) $\begin{equation} n_t= {\rm div}(\mu_n (\nabla n + n \varepsilon ) - n v ),\label{eq:lim-eqnp1} \end{equation}$
(1.11) $\begin{equation} p_t= {\rm div}(\mu_p (\nabla p - p \varepsilon ) - p v ),\label{eq:lim-eqnp2} \end{equation} $
(1.12) $\begin{equation} 0 =n - p -D(x), \label{eq:lim-eqnp3} \end{equation} $
(1.13) $\begin{equation} \rho v_t +\rho v\cdot \nabla v +A\nabla(\rho^2) -\mu \triangle v = \nabla{\rm div}v-(n - p ) \varepsilon, \label{eq:lim-eqnp4} \end{equation} $
(1.14) $\begin{equation} \rho_t+{\rm div}(\rho v)=0,\label{eq:lim-eqnp51} \end{equation} $
(1.15) $\begin{equation} (n,p,\rho,v)(x,0)=(n_0,p_0,\rho_0,v_0)(x),\label{eq:lim-eqnp7} \end{equation}$
其中$n_0(x)$ , $p_0(x)$ , $\rho_0(x)$ ,$v_0(x)$ 为光滑函数, 并且满足下面的条件
(1.16) $\begin{equation} n_0-p_0-D(x)=0.\label{eq:lim-eqnp8} \end{equation} $
据作者所知, 关于此系统的一些结论. 在掺杂函数光滑的假设下, Li[3 ] 证明了不可压电解液中电扩散方程的拟中性极限. Wang等[4 ] 研究了三维空间下带有不同迁移率电解液中不可压电扩散模型的拟中性极限和边界层问题. Wang等[5 ] 研究了电解液中不可压电扩散模型的初始层问题. Yang等[6 ] 研究了一般初值下量子 Navier-Stokes-Poisson 方程到不可压 Euler方程的收敛性问题. Liu等[7 ] 研究了三维空间中 Prandtl 非正定性. 拟中性问题被国内外专家学者广泛关注,并且有很多有意义的结论, 比如文献[8 ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ -19 ] 以及其参考文献.
该文主要研究迁移率互异的不可压PNPNS(1.1)-(1.6)的拟中性极限,并假设两个迁移率$\mu_n$ 与$\mu_p$ 的差适当小. 该文区别于不可压PNPNS 系统主要在于方程(1.5), 由于方程(1.5)中的散度不再是0, 这给能量估计带来很多困难, 比如不等式(3.21)的项. 幸运的是,通过借助不等式(3.22)-(3.23), 这些困难可以被很好的解决. 另外, 随着能量估计变得更加复杂, 该文中引入的两个$\lambda$ -加权的Lyapunov 函数(1.19)-(1.20)也变得复杂.
该文通过奇异摄动理论中的渐近匹配展开和加权的能量估计证明该文的结论. 该文的一个技巧是应用$|\mu_n-\mu_p|$ 的小性, 因为这样能保证系统是严格的抛物- 椭圆型系统, 可以直接应用相关结论. 为证明该文结论, 引入Gronwall型熵积分不等式
(1.17) $ \begin{matrix} \Gamma^\lambda(t)+\int_0^tG^\lambda(s){\rm d}s & \leq & M{\tilde{\Gamma}^\lambda}(t=0)+M\int_0^t(\Gamma^\lambda(s)+(\Gamma^\lambda(s))^2){\rm d}s\nonumber\\ && +M(\Gamma^\lambda(t))^2 +M\int_0^t\Gamma^\lambda(s)G^\lambda(s){\rm d}s+M\lambda,\label{eq:txg3}\end{matrix}$
(1.18) $\begin{matrix} \tilde{\Gamma^\lambda}(0)&=&\parallel(\bar{z}^\lambda,\bar{v}^\lambda,\bar{\rho}^\lambda,\bar{z}_t^\lambda,\bar{v}_t^\lambda,\bar{\rho}_t^\lambda,\nabla\bar{z}^\lambda,\nabla\bar{v}^\lambda,\nabla\bar{\rho}^\lambda,\nabla\bar{z}_t^\lambda,\nabla\bar{v}_t^\lambda,\nabla\bar{\rho}_t^\lambda)\parallel^2(t=0) \nonumber\\ &&+\lambda^2\parallel (\bar{E}^\lambda,\lambda{\rm div}\bar{E}^\lambda,\bar{E}_t^\lambda,\lambda{\rm div}\bar{E}_t^\lambda,\lambda\nabla{\rm div}\bar{E}^\lambda,\lambda\nabla{\rm div}\bar{E}_t^\lambda) \parallel^2(t=0).\label{eq:txg4} \end{matrix}$
另外, 为给出关于$\lambda$ 的一致先验估计, 通过引入两个$\lambda$ -加权的Lyapunov型函数
(1.19) $\begin{matrix} \Gamma^\lambda(t) &=&\parallel(\bar{z}^\lambda,\bar{v}^\lambda,\bar{E}^\lambda,\bar{\rho}^\lambda,\bar{z}_t^\lambda,\bar{v}_t^\lambda,\bar{\rho}_t^\lambda,\nabla\bar{z}^\lambda,\nabla\bar{v}^\lambda,{\rm div}\bar{v}^\lambda,\nabla\bar{\rho}^\lambda)\parallel^2\nonumber\\ &&+\parallel(\nabla\bar{z}_t^\lambda,\nabla\bar{v}_t^\lambda,\nabla\bar{\rho}_t^\lambda,\triangle\bar{z}^\lambda,\triangle\bar{v}^\lambda,\nabla{\rm div}\bar{v}^\lambda)\parallel^2\nonumber\\ &&+\lambda^2\parallel(\bar{E}^\lambda,\lambda{\rm div}\bar{E}^\lambda,\bar{E}_t^\lambda,\lambda{\rm div}\bar{E}_t^\lambda,\lambda\nabla{\rm div}\bar{E}^\lambda,\lambda\nabla{\rm div}\bar{E}_t^\lambda,\lambda\triangle{\rm div}\bar{E}^\lambda)\parallel^2\label{eq:tg1} \end{matrix} $
(1.20) $\begin{matrix} G^\lambda(t)&=&\parallel(\bar{\rho}^\lambda,\bar{E}^\lambda,\bar{E}_t^\lambda,\nabla\bar{z}^\lambda,\nabla\bar{v}^\lambda,{\rm div}\bar{v}^\lambda,\nabla\bar{\rho}^\lambda,\nabla\bar{z}_t^\lambda,\nabla\bar{v}_t^\lambda,{\rm div}\bar{v}_t^\lambda,\nabla\bar{\rho}_t^\lambda)\parallel^2\nonumber\\ &&+\parallel(\triangle\bar{z}^\lambda,\triangle\bar{v}^\lambda,\nabla{\rm div }\bar{v}^\lambda,\triangle\bar{z}_t^\lambda,\triangle\bar{v}_t^\lambda,\nabla{\rm div}\bar{v}_t^\lambda)\parallel^2\nonumber\\ &&+\lambda^2\parallel(\lambda{\rm div}\bar{E}^\lambda,\lambda{\rm div}\bar{E}_t^\lambda,\lambda\nabla{\rm div}\bar{E}^\lambda,\lambda\nabla{\rm div}\bar{E}_t^\lambda,\lambda\triangle{\rm div}\bar{E}^\lambda,\lambda\triangle{\rm div}\bar{E}^\lambda_t)\parallel^2,\label{eq:tg2} \end{matrix} $
其中$\bar{E}^\lambda$ , $\bar{v}^\lambda$ 和$\bar{\rho}^\lambda$ 由(2.1)式定义. 并且$\bar{z}^\lambda=\bar{n}^\lambda+\bar{p}^\lambda$ . 另外,该文应用$\epsilon,M,M(\epsilon)$ 表示独立于$\lambda$ 的正常数, 可能行于行之间不同.
2 误差方程和定理
(2.1) $\begin{equation} \bar{n}^\lambda=n^\lambda-n, \bar{p}^\lambda=p^\lambda-p, \bar{E}^\lambda=E^\lambda-\varepsilon, \bar{v}^\lambda=v^\lambda-v, \bar{\rho}^\lambda=\rho^\lambda-\rho, \label{eq:error-100} \end{equation}$
把(2.1)式代入系统(1.1)-(1.3),(1.5),(1.9)式. 并应用方程(1.10)-(1.14), 可得
(2.2) $\begin{equation} \bar{n}^\lambda_t={\rm div}(\mu_n(\nabla\bar{n}^\lambda+n\bar{E}^\lambda)+\bar{n}^\lambda(\bar{E}^\lambda+\varepsilon)-\bar{n}^\lambda(\bar{v}^\lambda+v)-n\bar{v}^\lambda),\label{eq:eqn1} \end{equation}$
(2.3) $\begin{equation} \bar{p}^\lambda_t={\rm div}(\mu_p(\nabla\bar{p}^\lambda-p\bar{E}^\lambda)-\bar{p}^\lambda(\bar{E}^\lambda+\varepsilon)-\bar{p}^\lambda(\bar{v}^\lambda+v)-p\bar{v}^\lambda),\label{eq:eqn2} \end{equation}$
(2.4) $\begin{equation} -\lambda^2{\rm div}\bar{E}^\lambda=\bar{n}^\lambda-\bar{p}^\lambda+\lambda^2{\rm div}\varepsilon,\label{eq:eqn3} \end{equation}$
(2.5) $\begin{matrix}\label{eqn4} &&(\bar{\rho}^\lambda+\rho)\bar{v}^\lambda_t+\bar{\rho} v_t+2A(\bar{\rho}^\lambda+\rho)\nabla\bar{\rho}^\lambda+2A\bar{\rho}^\lambda\nabla\rho+(\bar{\rho}^\lambda+\rho)(\bar{v}^\lambda+v)\nabla\bar{v}^\lambda\nonumber\\ &&+ (\bar{\rho}^\lambda+\rho)\bar{v}^\lambda\nabla v+\bar{\rho}^\lambda v \nabla v\nonumber\\ &=&\mu \Delta v+\nabla{\rm div} \bar{v}^\lambda+\lambda^2({\rm div}\bar{E}^\lambda+{\rm div}\varepsilon)(\bar{E}^\lambda+\varepsilon)-D \bar{E}^\lambda, \end{matrix}$
(2.6) $\begin{equation} \bar{\rho}^\lambda_t+{\rm div}((\rho+\bar{\rho}^\lambda)\bar{v}^\lambda+\rho^\lambda v)=0,\label{eq:eqn5} \end{equation}$
(2.7) $\begin{equation} (n^\lambda,p^\lambda,v^\lambda,\rho^\lambda)(x,0)=(n_0^\lambda(x),p_0^\lambda(x),v_0^\lambda(x),\rho_0^\lambda(x)).\label{eq:eqn6} \end{equation}$
(2.8) $\begin{equation} \bar{z}^\lambda = \bar{n}^\lambda +\bar{p}^\lambda, ~~ \bar{n}^\lambda =\frac{\bar{z}^\lambda - \lambda^2( {\rm div}\bar{E}^\lambda+{\rm div}\varepsilon)}{2}, ~~ \bar{p}^\lambda =\frac{\bar{z}^\lambda + \lambda^2( {\rm div}\bar{E}^\lambda+{\rm div}\varepsilon)}{2}\label{eq:trans1}, \end{equation}$
(2.9) $\begin{equation} \mu_n c\pm\mu_p d =\frac{\mu_n\pm\mu_p}{2}(c+d)+\frac{\mu_n\mp\mu_p}{2}(c-d),\label{eq:fab2} \end{equation}$
(2.10) $\begin{equation} A=\nabla \bar{z}^\lambda + D \bar{E}^\lambda - \lambda^2 \varepsilon {\rm div}\bar{E}^\lambda-\lambda^2(\bar{E}^\lambda+\varepsilon){\rm div}\varepsilon,\label{eq:dena} \end{equation}$
(2.11) $\begin{equation} B=-\lambda^2 \nabla({\rm div}\bar{E}^\lambda+{\rm div}\varepsilon)+ z\bar{E}^\lambda+\varepsilon \bar{z}^\lambda,\label{eq:denb} \end{equation}$
(2.12) $\begin{matrix} \bar{z}_t^\lambda& =&\frac{\mu_n + \mu_p}{2} {\rm div}A+ \frac{\mu_n - \mu_p}{2} {\rm div}B-{\rm div}(\bar{z}^\lambda v+z\bar{v}^\lambda)-\frac{\mu_n + \mu_p}{2} \lambda^2{\rm div}(\bar{E}^\lambda{\rm div}\bar{E}^\lambda)\nonumber\\ &&+\frac{\mu_n - \mu_p}{2} {\rm div}(\bar{z}^\lambda \bar{E}^\lambda)-{\rm div}(\bar{z}^\lambda \vec{v}^\lambda),\label{eq:eqz1} \end{matrix}$
(2.13) $\begin{matrix} -\lambda^2 {\rm div}E_t^\lambda &=&\frac{\mu_n + \mu_p}{2}{\rm div}B+\frac{\mu_n - \mu_p}{2}{\rm div}A+\lambda^2{\rm div}(v({\rm div}\bar{E}^\lambda+{\rm div}\varepsilon))\nonumber\\ &&+\lambda^2{\rm div}((\bar{v}^\lambda+v){\rm div}\varepsilon))-{\rm div}(D\bar{v}^\lambda)-\frac{\mu_n - \mu_p}{2} \lambda^2{\rm div}(\bar{E}^\lambda{\rm div}\bar{E}^\lambda)\nonumber\\ &&+\frac{\mu_n + \mu_p}{2} {\rm div}(\bar{z}^\lambda \bar{E}^\lambda)+\lambda^2{\rm div}(\bar{v}^\lambda {\rm div} \bar{E}^\lambda), \label{eq:eqz2} \end{matrix}$
(2.14) $\begin{matrix} \bar{\rho}^\lambda \bar{v}_t^\lambda +2A\bar{\rho}^\lambda\nabla\bar{\rho}^\lambda &=&-\rho \bar{v}^\lambda_t-2A\rho\nabla\bar{\rho}^\lambda-2A\bar{\rho}^\lambda \nabla\rho-\bar{\rho}^\lambda v_t-\bar{\rho}^\lambda v \nabla v\nonumber\\ &&-\rho \bar{v}^\lambda \nabla v-\rho v\nabla \bar{v}^\lambda +\lambda^2\bar{E}^\lambda{\rm div} \varepsilon+\lambda^2\varepsilon{\rm div}\bar{E}^\lambda+\lambda^2\varepsilon^\lambda {\rm div}\varepsilon\nonumber\\ &&-D\bar{E}^\lambda+\mu\Delta \bar{v}^\lambda+\nabla{\rm div}\bar{v}^\lambda-\rho \bar{v}^\lambda \nabla \bar{v}^\lambda-\bar{\rho}^\lambda v \nabla \bar{v}^\lambda-\bar{\rho}^\lambda \bar{v}^\lambda \nabla v\nonumber\\ &&+\lambda^2\bar{E}^\lambda{\rm div}\bar{E}^\lambda-\bar{\rho}^\lambda \bar{v}^\lambda\nabla \bar{v}^\lambda,\label{eq:eqz3} \end{matrix}$
(2.15) $\begin{equation} \bar{\rho}^\lambda_t+{\rm div}(\bar{\rho}^\lambda \bar{v}^\lambda)+{\rm div}(\rho \bar{v}^\lambda+v\bar{\rho}^\lambda)=0,\label{eq:eqz4} \end{equation}$
(2.16) $\begin{equation} (z^\lambda,v^\lambda,\rho^\lambda)(x,0)=(z_0^\lambda(x),v_0^\lambda(x),\rho_0^\lambda(x)). \label{eq:eqz6} \end{equation}$
形式上当$\lambda\rightarrow0$ 时, 有$(\bar{z}^\lambda, \bar{\rho}^\lambda, \bar{E}^\lambda, \bar{v}^\lambda )$ $\rightarrow$ $(z, \rho, \varepsilon, v)$ , 因此有极限系统
(2.17) $\begin{equation} z_t = \frac{\mu_n + \mu_p}{2} {\rm div}(\nabla z + D \varepsilon ) + \frac{\mu_n - \mu_p}{2} {\rm div} (3z \varepsilon ) -{\rm div}(3z v),\label{eq:lim-eqz1} \end{equation}$
(2.18) $\begin{equation} 0 = \frac{\mu_n + \mu_p}{2}{\rm div}(3z \varepsilon) + \frac{\mu_n - \mu_p}{2}{\rm div}(\nabla z + D \varepsilon )-{\rm div}(Dv), \label{eq:lim-eqz2} \end{equation}$
(2.19) $\begin{equation} \rho v_t +\rho v\cdot \nabla v +A\nabla(\rho^2) -\mu \triangle v = \nabla{\rm div}v-(n - p ) \varepsilon, \label{eq:lim-eqz3} \end{equation}$
(2.20) $\begin{equation} \rho_t+{\rm div}(\rho v)=0,\label{eq:lim-eqz4} \end{equation}$
(2.21) $\begin{equation} (n,p,\rho,v)(x,0)=(n_0,p_0,\rho_0,v_0)(x).\label{eq:lim-eqz5} \end{equation}$
易证系统(1.1)-(1.6)和系统(2.12)-(2.16), 以及极限系统(1.10)-(1.15)和(2.17)-(2.21)式的等价性,此处省略证明过程.
定理2.1 假设函数$(n^\lambda,p^\lambda,E^\lambda,v^\lambda,\rho^\lambda)$ 是系统(1.1)-(1.6)的局部解, 它们定义在 ${\Bbb T}^2\times[0,T^*)$ 上, 其中$T^*: 0<T^*<\infty$ 是极限系统(1.10)-(1.15)的局部光滑解$(n,p,\varepsilon,v,\rho)$ 的最大时间存在区间, 并且$n+p>k_0,0<\underline{\rho}<\rho<\overline{\rho}$ , 其中$k_0,\underline{\rho}<\rho<\overline{\rho}$ 是正常数, 同时假设存在正常数$M_1$ 使得
(2.22) $\begin{equation} \sum_{i=0}^3\parallel\partial_t^i(\bar{z}^\lambda,\lambda\bar{E}^\lambda,\lambda^2{\rm div}\bar{E}^\lambda,\bar{v}^\lambda,\bar{\rho}^\lambda)\parallel^2(t=0)\leq M_1\lambda \label{eq:ass1} \end{equation}$
(2.23) $\begin{equation} \sum_{i=0}^3\parallel\partial_t^i(\nabla\bar{z}^\lambda,\lambda{\rm div}\bar{E}^\lambda,\lambda^2\nabla{\rm div}\bar{E}^\lambda,\nabla\bar{v}^\lambda,\nabla\bar{\rho}^\lambda)\parallel^2(t=0)\leq M_1\lambda \label{eq:ass2} \end{equation}$
成立. 那么系统(2.12)-(2.16)存在唯一光滑解 $(\bar{z}^\lambda,\bar{p}^\lambda,\bar{E}^\lambda,\bar{v}^\lambda,\bar{\rho}^\lambda)$ 满足对任意的$T\in(0,T^*)$ 有正常数$M,\delta<1$ 和 $\lambda_0\ll1$ , 使得对任意的 $\lambda\in(0,\lambda_0]$ 有
(2.24) $\begin{equation} \sup_{0\leq t\leq T}(\parallel(\bar{z}^\lambda,\bar{v}^\lambda,\lambda\bar{E}^\lambda_t)\parallel^2_{H^2}+\parallel(\bar{\rho}^\lambda,\bar{z}^\lambda_t,\bar{v}^\lambda_t,\bar{\rho}^\lambda_t)\parallel^2_{H^1}+\parallel\bar{E}^\lambda\parallel^2_{H^3})\leq M \lambda^{1-\delta}. \label{eq:reus1} \end{equation}$
3 能量估计
为了简化符号, 引入$\lambda$ -加权Sobolev模如下
(3.1) $\begin{equation} \mid\mid\mid W\mid\mid\mid^2=\parallel(\bar{z}^\lambda,\bar{v}^\lambda,\lambda\bar{E}_t^\lambda)\parallel^2_{H^2}+\parallel (\bar{\rho}^\lambda,\bar{z}_t^\lambda,\bar{v}_t^\lambda,\bar{\rho}_t^\lambda) \parallel^2_{H^1}+\parallel\bar{E}^\lambda\parallel^2_{H^3}.~~~~\label{eq:tg3} \end{equation}$
(3.2) $\begin{matrix} &&\parallel(\bar{z}^\lambda,\lambda \bar{E}^\lambda,\lambda^2{\rm div}\bar{E}^\lambda,\bar{v}^\lambda,\bar{\rho}^\lambda)\parallel^2\nonumber\\ &&+M \int_0^t\parallel (\nabla \bar{z}^\lambda, \bar{E}^\lambda,\lambda {\rm div}\bar{E}^\lambda, \lambda^2\nabla {\rm div}\bar{E}^\lambda,\nabla \bar{v}^\lambda,{\rm div}\bar{v}^\lambda,\bar{\rho}^\lambda) \parallel^2{\rm d}t\nonumber\\ &\leq&\parallel(\bar{z}^\lambda,\lambda \bar{E}^\lambda,\lambda^2{\rm div}\bar{E}^\lambda,\bar{v}^\lambda,\bar{\rho}^\lambda)\parallel^2(t=0)\nonumber\\ &&+M\int_0^t\parallel ( \bar{z}^\lambda,\bar{v}^\lambda,\bar{\rho}^\lambda) \parallel^2{\rm d}t+M\int_0^t \mid\mid\mid W\mid\mid\mid^4(s){\rm d}s+M\lambda.\label{eq:eq-zevest1} \end{matrix}$
证 在系统(2.12)-(2.14)两端分别乘以$\bar{z}^\lambda,-\bar{\phi}^\lambda,\bar{v}^\lambda$ , 另外再在方程(2.13)两端乘以$\lambda^2 {\rm div}\bar{E}^\lambda$ , 再关于$x$ 在$ {\Bbb T}^3$ 上积分, 可得
(3.3) $\begin{matrix} &&\frac{\rm d}{{\rm d}t}\parallel(\bar{z}^\lambda,\lambda \bar{E}^\lambda,\lambda^2{\rm div}\bar{E}^\lambda,\bar{v}^\lambda,\bar{\rho}^\lambda)\parallel^2\nonumber \\ & &+k\parallel(\nabla \bar{z}^\lambda, \bar{E}^\lambda,\lambda {\rm div}\bar{E}^\lambda, \lambda^2\nabla {\rm div}\bar{E}^\lambda,\nabla \bar{v}^\lambda,{\rm div}\bar{v}^\lambda,\bar{\rho}^\lambda)\parallel^2\nonumber\\ & \leq & M\parallel(\bar{z}^\lambda,\bar{v}^\lambda,\bar{\rho}^\lambda)\parallel^2+M\mid\mid\mid W\mid\mid\mid^4+M\lambda, \label{eq:eq-zxy1} \end{matrix}$
对不等式(3.3)关于$t$ 在$[t]$ 上积分,可以得到不等式(3.2). 证毕
注3.1 定理3.1与定理3.2的证明类似, 此处省略该定理的证明.
(3.4) $\begin{matrix} &&\parallel(\bar{z}_t^\lambda,\lambda \bar{E}_t^\lambda,\lambda^2{\rm div}\bar{E}_t^\lambda,\bar{v}_t^\lambda,\bar{\rho}_t^\lambda)\parallel^2\nonumber \\ &&+M \int_0^t\parallel (\nabla \bar{z}_t^\lambda, \bar{E}_t^\lambda,\lambda {\rm div}\bar{E}_t^\lambda, \lambda^2\nabla {\rm div}\bar{E}_t^\lambda,\nabla \bar{v}_t^\lambda,{\rm div}\bar{v}_t^\lambda) \parallel^2{\rm d}t\nonumber \\ &\leq&\parallel(\bar{z}_t^\lambda,\lambda \bar{E}_t^\lambda,\lambda^2{\rm div}\bar{E}_t^\lambda,\bar{v}_t^\lambda,\bar{\rho}_t^\lambda)\parallel^2(t=0)\nonumber \\ &&+M\int_0^t\parallel ( \bar{z}^\lambda,\bar{z}_t^\lambda,\bar{v}^\lambda,\nabla\bar{v}^\lambda,{\rm div}\bar{v}^\lambda,\bar{v}_t^\lambda,\bar{\rho}^\lambda,\nabla\bar{\rho}^\lambda,\bar{\rho}_t^\lambda,\bar{E}^\lambda,{\rm div}\bar{E}^\lambda) \parallel^2{\rm d}t\nonumber \\ &&+M\int_0^t (\mid\mid\mid W\mid\mid\mid^4(s)+\mid\mid\mid W\mid\mid\mid^2G^\lambda(t)){\rm d}s+M\lambda.\label{eq:eq-zevest2} \end{matrix}$
证 对方程(2.12)关于$t$ 求导, 在方程两端乘以$\bar{z}^\lambda_t$ , 然后关于$x$ 在区间${\Bbb T}^3$ 上积分, 应用Green公式, 有
(3.5) $\begin{matrix} &&\frac{1}{2}\frac{\rm d}{{\rm d}t}\parallel \bar{z}_t^\lambda \parallel^2+\frac{\mu_n+\mu_p}{2}\parallel \nabla \bar{z}_t^\lambda\parallel^2\nonumber \\ &=&-\frac{\mu_n+\mu_p}{2}\int_{{\Bbb T}^3} (A-\nabla \bar{z}^\lambda)_t\nabla \bar{z}_t^\lambda {\rm d}x-\frac{\mu_n-\mu_p}{2}\int_{{\Bbb T}^3} B_t\nabla \bar{z}^\lambda_{t}{\rm d}x+\int_{{\Bbb T}^3} (z\bar{v}^\lambda+\bar{z}^\lambda v)_t\nabla \bar{z}^\lambda_t{\rm d}x\nonumber \\ &&-\frac{\mu_n-\mu_p}{2}\int_{{\Bbb T}^3}(\bar{z}^\lambda \bar{E}^\lambda)_t\nabla \bar{z}^\lambda_t{\rm d}x+\lambda^2\frac{\mu_n+\mu_p}{2}\int_{{\Bbb T}^3}(\bar{E}^\lambda{\rm div}\bar{E}^\lambda)_t\nabla \bar{z}^\lambda_t{\rm d}x+\int_{{\Bbb T}^3}(\bar{z}^\lambda \bar{v}^\lambda)_t\nabla \bar{z}^\lambda_t {\rm d}x. \label{eq:eq-zest8} \end{matrix}$
估计方程(3.5)右端各项. 对于线性项, 应用Cauchy-Schwarz 不等式, Sobolev嵌入定理, 线性项可以被(3.6)式控制
(3.6) $\begin{matrix} && \epsilon\frac{\mu_n+\mu_p}{2}\parallel\nabla \bar{z}^\lambda_{t}\parallel^2+\epsilon\parallel (\bar{E}^\lambda,\bar{E}^\lambda_{t},{\rm div}\bar{E}^\lambda,\bar{z}^\lambda,\bar{z}^\lambda_{t},\bar{v}^\lambda,\bar{v}^\lambda_{t}) \parallel^2\nonumber\\ && +M(\epsilon)\lambda^4\parallel ({\rm div}\bar{E}^\lambda_{t},\nabla{\rm div}\bar{E}^\lambda_{t}) \parallel^2+M\lambda.\label{eq:eq-zest9} \end{matrix}$
(3.7) $ \begin{matrix} &&\int_{{\Bbb T}^3} ({\bar{z}^\lambda \bar{v}^\lambda})_t\nabla \bar{z}^\lambda_{t} {\rm d}x\nonumber \\ & \leq & M(\epsilon)\parallel \bar{v}^\lambda_{t}\bar{z}^\lambda \parallel^2+M(\epsilon)\parallel \bar{v}^\lambda\bar{z}_t^\lambda \parallel^2+\epsilon\parallel \nabla \bar{z}^\lambda_{t} \parallel^2\nonumber \\ & \leq & M(\epsilon) \parallel \bar{v}^\lambda_{t} \parallel^2\parallel \bar{z}^\lambda\parallel_{H^2}^2+M(\epsilon) \parallel \bar{z}^\lambda_{t} \parallel^2\parallel \bar{v}^\lambda\parallel_{H^2}^2+\epsilon\parallel \nabla \bar{z}^\lambda_{t} \parallel^2\nonumber \\ &\leq& \epsilon\parallel \nabla \bar{z}^\lambda_{t} \parallel^2+M\mid\mid\mid W\mid\mid\mid^4,\label{eq:eq-zest10} \end{matrix}$
(3.8) $\begin{matrix} &&-\frac{\mu_n-\mu_p}{2}\int_{{\Bbb T}^3} ({\bar{z}^\lambda\bar{E}^\lambda})_t\nabla \bar{z}^\lambda_t {\rm d}x\nonumber \\ &=&-\frac{\mu_n-\mu_p}{2}\int_{{\Bbb T}^3} ({\bar{z}^\lambda_t\bar{E}^\lambda+\bar{z}^\lambda\bar{E}_t^\lambda})\nabla \bar{z}^\lambda_t {\rm d}x\nonumber \\ &\leq & M(\epsilon)(\parallel\bar{E}^\lambda_{t}\bar{z}^\lambda\parallel^2+\parallel \bar{E}^\lambda\bar{z}^\lambda_{t} \parallel^2)+\epsilon\frac{\mu_n-\mu_p}{2}\parallel \nabla \bar{z}^\lambda_{t} \parallel^2\nonumber\\ & \leq& M(\epsilon)( \parallel \bar{z}^\lambda\parallel_{H^2}^2\parallel \bar{E}^\lambda_{t} \parallel^2+\parallel \bar{z}^\lambda_{t} \parallel_{H^1}^2\parallel \bar{E}^\lambda\parallel_{H^1}^2)+\epsilon\frac{\mu_n-\mu_p}{2}\parallel \nabla \bar{z}^\lambda_{t} \parallel^2\nonumber \\ &\leq&\epsilon\frac{\mu_n-\mu_p}{2}\parallel \nabla \bar{z}^\lambda_{t} \parallel^2+M\mid\mid\mid W\mid\mid\mid^4,\label{eq:eq-zest11} \end{matrix}$
(3.9) $ \begin{matrix} &&\lambda^2\frac{\mu_n+\mu_p}{2}\int_{{\Bbb T}^3} ({\bar{E}^\lambda{\rm div}\bar{E}^\lambda})_t\nabla \bar{z}^\lambda_{t} {\rm d}x\nonumber \\ &=&\lambda^2\frac{\mu_n+\mu_p}{2}\int_{{\Bbb T}^3}({\bar{E}^\lambda_{t}{\rm div}\bar{E}^\lambda}+{\bar{E}^\lambda{\rm div}\bar{E}^\lambda_{t}})\nabla \bar{z}^\lambda_{t} {\rm d}x \nonumber \\ & \leq &\lambda^4M(\epsilon)( \parallel \bar{E}^\lambda_{t}{\rm div}\bar{E}^\lambda \parallel^2+\parallel \bar{E}^\lambda{\rm div}\bar{E}^\lambda_{t} \parallel^2)+\epsilon\frac{\mu_n+\mu_p}{2}\parallel \nabla \bar{z}^\lambda_{t} \parallel^2\nonumber \\ & \leq &\lambda^4M(\epsilon)(\parallel \bar{E}^\lambda \parallel_{H^2}^2\parallel {\rm div}\bar{E}^\lambda_{t} \parallel^2+\parallel \bar{E}^\lambda_{t} \parallel_{H^1}^2\parallel {\rm div}\bar{E}^\lambda \parallel_{H^1}^2)+\epsilon\frac{\mu_n+\mu_p}{2}\parallel \nabla \bar{z}^\lambda_{t} \parallel^2\nonumber \\ &\leq&\epsilon\frac{\mu_n+\mu_p}{2}\parallel \nabla \bar{z}^\lambda_{t} \parallel^2+M\mid\mid\mid W\mid\mid\mid^4.\label{eq:eq-zest12} \end{matrix}$
(3.10) $\begin{matrix} && \frac{\rm d}{{\rm d}t}\parallel \bar{z}^\lambda_{t} \parallel^2+k\parallel \nabla \bar{z}^\lambda_{t}\parallel^2\nonumber \\ & \leq &M\parallel(\bar{z}^\lambda,\bar{z}^\lambda_{t},\bar{E}^\lambda,\bar{E}^\lambda_{t},{\rm div}\bar{E}^\lambda,\bar{v}^\lambda,\bar{v}^\lambda_{t})\parallel^2 +M\lambda^4\parallel({\rm div}\bar{E}^\lambda_{t},\nabla {\rm div}\bar{E}^\lambda_{t})\parallel^2\nonumber \\ && +M\mid\mid\mid W\mid\mid\mid^4+M\lambda.\label{eq:eq-zest17} \end{matrix}$
对方程(2.13)关于$t$ 求导, 两端乘以$\bar{\phi}^\lambda_{t}$ , 再关于$x$ 在区间${\Bbb T}^3$ 积分, 可得
(3.11) $\begin{matrix} &&\frac{\lambda^2}{2}\frac{\rm d}{{\rm d}t}\parallel \bar{E}^\lambda_{t} \parallel^2+\lambda^2\parallel {\rm div} \bar{E}^\lambda_{t}\parallel^2+\frac{\mu_n+\mu_p}{2}\int_{{\Bbb T}^3} z |\bar{E}^\lambda_{t}|^2{\rm d}x\nonumber \\ & =&\frac{\mu_n+\mu_p}{2}\int_{{\Bbb T}^3}B_t\bar{E}^\lambda_{t}{\rm d}x+\frac{\mu_n+\mu_p}{2}\lambda^2\int_{{\Bbb T}^3} z |{\rm div}\bar{E}^\lambda_{t}|^2{\rm d}x-\frac{\mu_n+\mu_p}{2}\int_{{\Bbb T}^3} z_t\bar{E}^\lambda \bar{E}^\lambda_{t}{\rm d}x\nonumber \\ && -\frac{\mu_n-\mu_p}{2}\int_{{\Bbb T}^3}A_t\bar{E}^\lambda_{t}{\rm d}x+\lambda^2\int_{{\Bbb T}^3}(v({\rm div}\bar{E}^\lambda+{\rm div}\varepsilon))_t\bar{E}^\lambda_{t}{\rm d}x\nonumber \\ && -\lambda^2\int_{{\Bbb T}^3} ((\bar{v}^\lambda+v){\rm div}\varepsilon)_t \bar{E}^\lambda_{t}{\rm d}x +\int_{{\Bbb T}^3} D\bar{v}^\lambda_t \bar{E}^\lambda_{t}{\rm d}x-\frac{\mu_n+\mu_p}{2}\int_{{\Bbb T}^3} {(\bar{z}^\lambda\bar{E}^\lambda)_t}\bar{E}^\lambda_{t} {\rm d}x\nonumber \\ &&-\lambda^2\int_{{\Bbb T}^3} (\bar{v}^\lambda{\rm div}\bar{E}^\lambda)_t\bar{E}^\lambda_{t} {\rm d}x +\lambda^2\frac{\mu_n-\mu_p}{2}\int_{{\Bbb T}^3} {(\bar{E}^\lambda{\rm div}\bar{E}^\lambda)_t}\bar{E}^\lambda_{t} {\rm d}x. \label{eq:eq-eest7} \end{matrix}$
估计(3.11)式右端各项. 对于线性项, 应用Cauchy-Schwarz不等式, Sobolev嵌入定理, $\bar{E}^\lambda_{t}=-\nabla\bar{\phi}^\lambda_{t}$ , 线性项可以被(3.12)控制
(3.12) $\begin{equation} \epsilon\parallel (\bar{E}^\lambda,\bar{E}^\lambda_{t},\bar{z}^\lambda,\bar{z}^\lambda_{t},\bar{v}^\lambda,\bar{v}^\lambda_{t}) \parallel^2 +M(\epsilon)\lambda^4 \parallel ({\rm div}\bar{E}^\lambda,{\rm div}\bar{E}^\lambda_{t}) \parallel^2+M\lambda,\label{eq:eq-eest8} \end{equation}$
对于非线性项, 应用Cauchy-Schwarz不等式和Sobolev嵌入不等式, 有
(3.13) $\begin{matrix} &&-\frac{\mu_n+\mu_p}{2}\int_{{\Bbb T}^3} {(\bar{z}^\lambda\bar{E}^\lambda)_t}\bar{E}^\lambda_{t} {\rm d}x\nonumber \\ &=&-\frac{\mu_n+\mu_p}{2}\int_{{\Bbb T}^3} {(\bar{z}^\lambda_{t}\bar{E}^\lambda+\bar{z}^\lambda\bar{E}^\lambda_{t})}\bar{E}^\lambda_{t} {\rm d}x\nonumber \\ & \leq & M(\epsilon)\parallel \bar{E}^\lambda_{t} \bar{z}^\lambda \parallel^2+ M(\epsilon)\parallel \bar{E}^\lambda \bar{z}^\lambda_{t} \parallel^2+\epsilon\parallel \bar{E}^\lambda_{t} \parallel^2\nonumber \\ & \leq &M(\epsilon)\parallel \bar{z}^\lambda \parallel_{H^2}^2 \parallel \bar{E}^\lambda_{t} \parallel^2+M(\epsilon)\parallel \bar{E}^\lambda \parallel_{H^1}^2 \parallel \bar{z}^\lambda_{t} \parallel_{H^1}^2+\epsilon\parallel \bar{E}^\lambda_{t} \parallel^2\nonumber \\ &\leq&\epsilon\parallel \bar{E}^\lambda_{t} \parallel^2+M\mid\mid\mid W\mid\mid\mid^4+M\mid\mid\mid W\mid\mid\mid^2G^\lambda(t),\label{eq:eq-eest9} \end{matrix}$
(3.14) $\begin{matrix} && \lambda^2\frac{\mu_n-\mu_p}{2}\int_{{\Bbb T}^3}{(\bar{E}^\lambda{\rm div}\bar{E}^\lambda)_t}\bar{E}^\lambda_{t} {\rm d}x\nonumber \\ & =&\lambda^2\frac{\mu_n-\mu_p}{2}\int_{{\Bbb T}^3}{\bar{E}^\lambda_{t}{\rm div}\bar{E}^\lambda}\bar{E}^\lambda_{t} {\rm d}x+\frac{\mu_n-\mu_p}{2}\int_{{\Bbb T}^3} \lambda^2{\bar{E}^\lambda{\rm div}\bar{E}^\lambda_{t}}\bar{E}^\lambda_{t} {\rm d}x\nonumber \\ &\leq& M(\epsilon)\lambda^4\parallel \bar{E}^\lambda_{t} {\rm div}\bar{E}^\lambda\parallel^2+M(\epsilon)\lambda^4\parallel \bar{E}^\lambda {\rm div}\bar{E}^\lambda_{t}\parallel^2+\epsilon(\frac{\mu_n-\mu_p}{2})^2\parallel \bar{E}^\lambda_{t} \parallel^2\nonumber \\ & \leq& M\lambda^4\parallel \bar{E}^\lambda\parallel_{H^2}^2 \parallel {\rm div}\bar{E}^\lambda_{t}\parallel^2+M\lambda^4\parallel \bar{E}^\lambda_{t}\parallel_{H^1}^2 \parallel {\rm div}\bar{E}^\lambda\parallel_{H^1}^2+\epsilon(\frac{\mu_n-\mu_p}{2})^2\parallel \bar{E}^\lambda_{t} \parallel^2\nonumber \\ &\leq&\epsilon\parallel \bar{E}^\lambda_{t} \parallel^2+M\mid\mid\mid W\mid\mid\mid^4,\label{eq:eq-eest10} \end{matrix}$
(3.15) $\begin{matrix} && -\lambda^2 \int_{{\Bbb T}^3}({\rm div}\bar{E}^\lambda \bar{v}^\lambda)_t\bar{E}^\lambda_{t} {\rm d}x\nonumber \\ & =&-\lambda^2 \int_{{\Bbb T}^3} {\rm div}\bar{E}^\lambda \bar{v}^\lambda_{t}\bar{E}^\lambda_{t} {\rm d}x-\lambda^2 \int_{{\Bbb T}^3} {\rm div}\bar{E}^\lambda_{t} \bar{v}^\lambda\bar{E}^\lambda_{t} {\rm d}x\nonumber \\ & \leq& M(\epsilon)\lambda^4\parallel \bar{v}^\lambda_{t} {\rm div}\bar{E}^\lambda\parallel^2+M(\epsilon)\lambda^4\parallel \bar{v}^\lambda {\rm div}\bar{E}^\lambda_{t}\parallel^2+\epsilon\parallel \bar{E}^\lambda_{t} \parallel^2\nonumber \\ & \leq& M\lambda^4\parallel \bar{v}^\lambda_{t}\parallel_{H^1}^2 \parallel {\rm div}\bar{E}^\lambda\parallel_{H^1}^2+M\lambda^4\parallel \bar{v}^\lambda\parallel_{H^2}^2 \parallel {\rm div}\bar{E}^\lambda_{t}\parallel^2+\epsilon\parallel \bar{E}^\lambda_{t} \parallel^2\nonumber\\ & \leq&\epsilon\parallel \bar{E}^\lambda_{t} \parallel^2+M\mid\mid\mid W\mid\mid\mid^4.\label{eq:eq-eest11} \end{matrix}$
那么, 由不等式(3.11)-(3.15), 并限制$\lambda$ 足够小, 有
(3.16) $\begin{matrix} && \lambda^2 \frac{\rm d}{{\rm d}t}\parallel \bar{E}^\lambda_{t} \parallel^2+k\parallel (\bar{E}^\lambda_{t},\lambda {\rm div} \bar{E}^\lambda_{t})\parallel^2\nonumber \\ & \leq & M(\epsilon)\parallel(\bar{z}^\lambda,\bar{z}^\lambda_{t},\nabla \bar{z}^\lambda_{t},\bar{E}^\lambda,{\rm div}\bar{E}^\lambda,\bar{v}^\lambda,\bar{v}^\lambda_{t})\parallel^2+M\mid\mid\mid W\mid\mid\mid^4\nonumber \\ && +M\mid\mid\mid W\mid\mid\mid^2G^\lambda(t)+M\lambda.\label{eq:eq-eest13} \end{matrix}$
对(2.13)式关于$t$ 求导, 两边乘以$\lambda^2{\rm div}\bar{E}^\lambda_{t}$ , 关于$x$ 在区间${{\Bbb T}^3}$ 上积分, 可得
(3.17) $\begin{matrix} &&\frac{\lambda^4}{2}\frac{\rm d}{{\rm d}t}\parallel {\rm div}\bar{E}^\lambda_{t} \parallel^2+\lambda^4\parallel \nabla{\rm div} \bar{E}^\lambda_{t}\parallel^2+\frac{\mu_n+\mu_p}{2}\lambda^2\int_{{\Bbb T}^3} z|{\rm div}\bar{E}^\lambda_{t}|^2{\rm d}x\nonumber \\ &=&\frac{\mu_n+\mu_p}{2}\int_{{\Bbb T}^3}B_t\nabla{\rm div}\bar{E}^\lambda_{t}{\rm d}x+\frac{\mu_n+\mu_p}{2}\lambda^2\int_{{\Bbb T}^3} z |{\rm div}\bar{E}^\lambda_{t}|^2{\rm d}x-\frac{\mu_n+\mu_p}{2}\int_{{\Bbb T}^3} z_t\bar{E}^\lambda\nabla{\rm div} \bar{E}^\lambda_{t}{\rm d}x\nonumber \\ &&-\frac{\mu_n-\mu_p}{2}\int_{{\Bbb T}^3}A_t\nabla{\rm div}\bar{E}^\lambda_{t}{\rm d}x+\lambda^2\int_{{\Bbb T}^3}(v({\rm div}\bar{E}^\lambda+{\rm div}\varepsilon))_t\nabla{\rm div}\bar{E}^\lambda_{t}{\rm d}x\nonumber \\ &&-\lambda^2\int_{{\Bbb T}^3} ((\bar{v}^\lambda+v){\rm div}\varepsilon)_t \nabla{\rm div}\bar{E}^\lambda_{t}{\rm d}x +\int_{{\Bbb T}^3} D\bar{v}^\lambda_t \nabla{\rm div}\bar{E}^\lambda_{t}{\rm d}x-\frac{\mu_n+\mu_p}{2}\int_{{\Bbb T}^3} {(\bar{z}^\lambda\bar{E}^\lambda)_t}\nabla{\rm div}\bar{E}^\lambda_{t} {\rm d}x\nonumber \\ &&-\lambda^2\int_{{\Bbb T}^3} (\bar{v}^\lambda{\rm div}\bar{E}^\lambda)_t\nabla{\rm div}\bar{E}^\lambda_{t} {\rm d}x +\lambda^2\frac{\mu_n-\mu_p}{2}\int_{{\Bbb T}^3} {(\bar{E}^\lambda{\rm div}\bar{E}^\lambda)_t}\nabla{\rm div}\bar{E}^\lambda_{t} {\rm d}x. \label{eq:eq-ee1} \end{matrix}$
由于方程(3.17)与方程(3.11)完全类似, 因此, 同理可得
(3.18) $\begin{matrix} &&\lambda^4 \frac{\rm d}{{\rm d}t}\parallel {\rm div}\bar{E}^\lambda_{t} \parallel^2+k\lambda^2\parallel ({\rm div}\bar{E}^\lambda_{t},\lambda \nabla{\rm div} \bar{E}^\lambda_{t})\parallel^2\nonumber \\ &\leq & M(\epsilon)\parallel(\bar{z}^\lambda,\bar{z}^\lambda_{t},\nabla \bar{z}^\lambda_{t},\bar{E}^\lambda,\bar{v}^\lambda,\bar{v}^\lambda_{t})\parallel^2+M\lambda^2\parallel(\bar{E}^\lambda,\bar{E}^\lambda_{t},{\rm div}\bar{E}^\lambda)\parallel^2\nonumber \\ && +M\mid\mid\mid W\mid\mid\mid^4+M\mid\mid\mid W\mid\mid\mid^2G^\lambda(t)+M\lambda.\label{eq:eq-ee7} \end{matrix}$
对方程(2.14)关于$t$ 求导, 乘以$\bar{v}^\lambda_{t}$ ,在区间${{\Bbb T}^3}$ 上关于$x$ 积分, 分部积分,有
(3.19) $\begin{matrix} &&\int_{{\Bbb T}^3}(\rho+\bar{\rho}^\lambda)\bar{v}^\lambda_{tt}\bar{v}^\lambda_t {\rm d}x+A\frac{\rm d}{{\rm d}t}\parallel\bar{\rho}^\lambda_t\parallel^2+\mu\parallel\nabla\bar{v}^\lambda_t\parallel^2+\parallel{\rm div}\bar{v}^\lambda_t\parallel^2\nonumber \\ & =&-\int_{{\Bbb T}^3}(\bar{\rho}^\lambda v_t+\bar{\rho}^\lambda v \nabla v+\rho\bar{v}^\lambda \nabla v+\rho v \nabla\bar{v}^\lambda+\rho_t\bar{v}^\lambda_t )_t\bar{v}^\lambda_t{\rm d}x\nonumber\\ && -\int_{{\Bbb T}^3}D\bar{E}^\lambda_t \bar{v}^\lambda_t {\rm d}x-\lambda^2\int_{{\Bbb T}^3}(\bar{E}^\lambda{\rm div}\varepsilon+\varepsilon{\rm div}\bar{E}^\lambda+\varepsilon{\rm div}\varepsilon)_t\bar{v}_t^\lambda {\rm d}x\nonumber \\ &&-2A\int_{{\Bbb T}^3}(\bar{\rho}^\lambda{\rm div}v+v \nabla\bar{\rho}^\lambda)_t\rho^\lambda_t{\rm d}x-2A\int_{{\Bbb T}^3}(\rho_t{\rm div}\bar{v}^\lambda+\bar{v}^\lambda \nabla\rho_t)\bar{\rho}^\lambda_t{\rm d}x\nonumber \\ && -2A\int_{{\Bbb T}^3}(\rho \bar{v}^\lambda \nabla\bar{v}^\lambda+\bar{\rho}^\lambda v \nabla\bar{v}^\lambda+\bar{\rho}^\lambda\bar{v}^\lambda\nabla\bar{v}^\lambda)_t \bar{v}_t^\lambda {\rm d}x\nonumber \\ &&-\int_{{\Bbb T}^3}((\bar{\rho}^\lambda\bar{v}^\lambda\nabla v)_t+\bar{\rho}_t^\lambda\bar{v}^\lambda_t+\lambda^2(\bar{E}^\lambda{\rm div}\bar{E}^\lambda)_t)\bar{v}^\lambda_t {\rm d}x\nonumber \\ && -2A\int_{{\Bbb T}^3}({\rm div}\bar{v}^\lambda\bar{\rho}^\lambda_t+\bar{v}^\lambda\nabla\bar{\rho}^\lambda_t)\bar{\rho}^\lambda_t{\rm d}x.\label{eq:eq-vest6} \end{matrix}$
估计方程(3.19)右端各项. 线性项可以被(3.20)式控制
(3.20) $\begin{matrix} &&\epsilon\parallel(\bar{\rho}^\lambda,\bar{\rho}^\lambda_t,\nabla\bar{\rho}^\lambda,\bar{v}^\lambda,\nabla\bar{v}^\lambda,\bar{v}^\lambda_t,\nabla\bar{v}^\lambda_t,{\rm div}\bar{v}^\lambda,\bar{E}^\lambda,\bar{E}^\lambda_t)\parallel^2\nonumber \\ && +M(\epsilon)\lambda^4\parallel (\bar{E}^\lambda,\bar{E}^\lambda_t,{\rm div}\bar{E}^\lambda,{\rm div}\bar{E}^\lambda_t) \parallel^2+M\lambda.~~~~\label{eq:eq-vest7} \end{matrix}$
对于非线性项, 应用Cauchy-Schwarz不等式, Sobolev嵌入不等式, 见文献[4 ], 有
(3.21) $\begin{matrix} &&-2A\int_{{\Bbb T}^3}(\rho_t{\rm div}\bar{v}^\lambda+\bar{v}^\lambda \nabla\rho_t)\bar{\rho}^\lambda_t{\rm d}x-2A\int_{{\Bbb T}^3}(\rho \bar{v}^\lambda \nabla\bar{v}^\lambda+\bar{\rho}^\lambda v \nabla\bar{v}^\lambda+\bar{\rho}^\lambda\bar{v}^\lambda\nabla\bar{v}^\lambda)_t \bar{v}_t^\lambda {\rm d}x\nonumber \\ &&-\int_{{\Bbb T}^3}((\bar{\rho}^\lambda\bar{v}^\lambda\nabla v)_t+\bar{\rho}_t^\lambda\bar{v}^\lambda_t+\lambda^2(\bar{E}^\lambda{\rm div}\bar{E}^\lambda)_t)\bar{v}^\lambda_t {\rm d}x-2A\int_{{\Bbb T}^3}({\rm div}\bar{v}^\lambda\bar{\rho}^\lambda_t+\bar{v}^\lambda\nabla\bar{\rho}^\lambda_t)\bar{\rho}^\lambda_t{\rm d}x\nonumber \\ &\leq& \epsilon \parallel(\bar{v}^\lambda_t,\nabla\bar{v}^\lambda_t,\bar{\rho}^\lambda_t)\parallel^2+M\parallel\bar{v}^\lambda\bar{v}^\lambda_t \parallel^2+M\parallel\bar{\rho}^\lambda\bar{v}^\lambda_t\parallel^2\nonumber \\ && +M\parallel\nabla\bar{v}^\lambda\bar{v}^\lambda_t\parallel^2+M\parallel\bar{\rho}^\lambda_t\nabla\bar{v}^\lambda\parallel^2+M\parallel\bar{v}^\lambda\bar{\rho}^\lambda_t\parallel^2+M\parallel\bar{\rho}^\lambda_t\bar{v}^\lambda_t\parallel^2\nonumber \\ && +\lambda^4M\parallel\bar{E}^\lambda_t{\rm div}\bar{E}^\lambda\parallel^2+M\lambda^4\parallel\bar{E}^\lambda{\rm div}\bar{E}^\lambda_t\parallel^2+M\parallel\bar{\rho}^\lambda_t{\rm div}\bar{v}^\lambda\parallel^2\nonumber\\ & \leq& \epsilon \parallel(\bar{v}^\lambda_t,\nabla\bar{v}^\lambda_t,\bar{\rho}^\lambda_t)\parallel^2+M\parallel\bar{v}^\lambda_t \parallel^2\parallel\bar{v}^\lambda\parallel^2_{H^2}+M\parallel\bar{v}^\lambda_t \parallel^2_{H^1}\parallel\bar{\rho}^\lambda\parallel^2_{H^1}\nonumber \\ && +M\parallel\bar{v}^\lambda_t \parallel^2_{H^1}\parallel\nabla\bar{v}^\lambda\parallel^2_{H^1}+M\parallel\bar{\rho}^\lambda_t \parallel^2_{H^1}\parallel\nabla\bar{v}^\lambda\parallel^2_{H^1}+M\parallel\bar{\rho}^\lambda_t \parallel^2\parallel\bar{v}^\lambda\parallel^2_{H^2}\nonumber \\ && +M\parallel\bar{v}^\lambda_t \parallel^2_{H^1}\parallel\bar{\rho}^\lambda_t\parallel^2_{H^1}+M\parallel\bar{\rho}^\lambda_t \parallel^2_{H^1}\parallel{\rm div}\bar{v}^\lambda\parallel^2_{H^1}\nonumber \\ && +M\lambda^4\parallel\bar{E}^\lambda_t \parallel^2_{H^1}\parallel{\rm div}\bar{E}^\lambda\parallel^2_{H^1}+M\lambda^4\parallel\bar{E}^\lambda \parallel^2_{H^2}\parallel{\rm div}\bar{E}^\lambda_t\parallel^2\nonumber\\ & \leq& \epsilon \parallel(\bar{v}^\lambda_t,\nabla\bar{v}^\lambda_t,\bar{\rho}^\lambda_t)\parallel^2+M\mid\mid\mid W\mid\mid\mid^4.\label{eq:eq-vest9} \end{matrix}$
联合不等式(3.19)-(3.21), 并应用不等式(3.22),有
(3.22) $\begin{equation} \int_{{\Bbb T}^3}(\rho+\bar{\rho}^\lambda)\bar{v}^\lambda_{tt}\bar{v}^\lambda_t {\rm d}x\geq \frac{\underline{\rho}}{2}\int_{{\Bbb T}^3}\bar{v}^\lambda_{tt}\bar{v}^\lambda_t {\rm d}x=\frac{\underline{\rho}}{4}\frac{\rm d}{{\rm d}t}\parallel\bar{v}^\lambda_t\parallel^2,\label{eq:rvt1} \end{equation}$
$\rho$ 和$\bar{\rho}^\lambda$ 满足不等式
(3.23) $\begin{equation} \frac{\underline{\rho}}{2}\leq \rho-|\bar{\rho}^\lambda|_{L^\infty}\leq\rho+\bar{\rho}^\lambda\leq\overline{\rho}+|\bar{\rho}^\lambda|_{L^\infty}\leq2\overline{\rho},\label{eq:rvt2} \end{equation} $
其中$\underline{\rho}$ 和$\overline{\rho}$ 分别是下确界和上确界, 因此可得
(3.24) $\begin{matrix} &&\epsilon\frac{\rm d}{{\rm d}t}\parallel(\bar{\rho}^\lambda_t,\bar{v}^\lambda_t) \parallel^2+k\parallel (\nabla\bar{v}^\lambda_t,{\rm div}\bar{v}^\lambda_t)\parallel^2 \nonumber \\ & \leq & M\parallel(\bar{v}^\lambda,\bar{v}^\lambda_t,\nabla\bar{v}^\lambda,{\rm div}\bar{v}^\lambda,\bar{\rho}^\lambda,\bar{\rho}^\lambda_t,\bar{E}^\lambda, \bar{E}^\lambda_t)\parallel^2\nonumber \\ && +M\lambda^4\parallel({\rm div}\bar{E}^\lambda,{\rm div} \bar{E}^\lambda_t)\parallel^2 +M\mid\mid\mid W\mid\mid\mid^4+M\lambda.\label{eq:eq-vest10} \end{matrix}$
联合$\delta_2$ (3.10) ,(3.15),(3.18)和(3.24)式, 在区间$[t]$ 上关于$t$ 积分, 取$\delta_2$ $>$ $0$ 足够小, 可得不等式(3.4). 证毕
定理3.3 估计$\parallel(\nabla \bar{z}^\lambda, \bar{E}^\lambda,\lambda {\rm div}\bar{E}^\lambda, \lambda^2\nabla {\rm div}\bar{E}^\lambda,\nabla \bar{v}^\lambda,{\rm div}\bar{v}^\lambda,\bar{\rho}^\lambda)\parallel^2$ ,有
(3.25) $\begin{matrix} &&k\parallel(\nabla \bar{z}^\lambda, \bar{E}^\lambda,\lambda {\rm div}\bar{E}^\lambda, \lambda^2\nabla {\rm div}\bar{E}^\lambda,\nabla \bar{v}^\lambda,{\rm div}\bar{v}^\lambda,\bar{\rho}^\lambda)\parallel^2\nonumber \\ & \leq& M\parallel(\bar{z}^\lambda,\bar{z}^\lambda_t,\bar{v}^\lambda,\bar{v}^\lambda_t,\bar{\rho}^\lambda,,\bar{\rho}^\lambda_t)\parallel^2+M\lambda^2\parallel(\bar{E}^\lambda,\bar{E}^\lambda_t,\lambda{\rm div}\bar{E}^\lambda,\lambda{\rm div}\bar{E}^\lambda_t)\parallel^2\nonumber \\ && +M\mid\mid\mid W\mid\mid\mid^4+M\lambda,\label{eq:lw1} \end{matrix}$
注3.2 不等式(3.25)可由不等式(3.3)和Green公式得到, 此处省略其证明.
(3.26) $\begin{matrix} &&\parallel(\nabla\bar{z}^\lambda,\lambda \bar{\rm div}{E}^\lambda,\lambda^2\nabla{\rm div}\bar{E}^\lambda,\bar\nabla{v}^\lambda,\bar\nabla{\rho}^\lambda)\parallel^2\nonumber \\ && +M \int_0^t\parallel (\triangle \bar{z}^\lambda, {\rm div}\bar{E}^\lambda,\lambda\nabla {\rm div}\bar{E}^\lambda, \lambda^2\triangle {\rm div}\bar{E}^\lambda,\triangle \bar{v}^\lambda,\nabla{\rm div}\bar{v}^\lambda,\nabla\bar{\rho}^\lambda) \parallel^2 dt\nonumber \\ & \leq&\parallel(\nabla\bar{z}^\lambda,\lambda \bar{\rm div}{E}^\lambda,\lambda^2\nabla{\rm div}\bar{E}^\lambda,\bar\nabla{v}^\lambda,\bar\nabla{\rho}^\lambda)\parallel^2(t=0)\nonumber \\ && +M\int_0^t\parallel ( \nabla\bar{z}^\lambda,\nabla\bar{v}^\lambda,\nabla\bar{\rho}^\lambda) \parallel^2{\rm d}t+M\int_0^t \mid\mid\mid W\mid\mid\mid^4(s){\rm d}s+M\lambda.\label{eq:eq-zevest3} \end{matrix}$
证 在方程(2.12)-(2.14)两边分别乘以$-\triangle\bar{z}^\lambda,-\triangle\bar{\phi}^\lambda,-\triangle\bar{v}^\lambda$ , 再在方程(2.13)两边乘以$\lambda^2 \triangle{\rm div}\bar{E}^\lambda$ , 应用$|\mu_n-\mu_p|$ 的小性, 在区间${\Bbb T}^3$ 上关于$x$ 积分, 分部积分, 应用Green公式, 有
(3.27) $\begin{matrix} &&\frac{\rm d}{{\rm d}t}\parallel(\nabla\bar{z}^\lambda,\lambda \bar{\rm div}{E}^\lambda,\lambda^2\nabla{\rm div}\bar{E}^\lambda,\bar\nabla{v}^\lambda,\bar\nabla{\rho}^\lambda)\parallel^2\nonumber \\ && +k\parallel(\triangle \bar{z}^\lambda, {\rm div}\bar{E}^\lambda,\lambda\nabla {\rm div}\bar{E}^\lambda, \lambda^2\triangle {\rm div}\bar{E}^\lambda,\triangle \bar{v}^\lambda,\nabla{\rm div}\bar{v}^\lambda,\nabla\bar{\rho}^\lambda)\parallel^2\nonumber \\ & \leq& M\parallel(\nabla\bar{z}^\lambda,\nabla\bar{v}^\lambda,\nabla\bar{\rho}^\lambda)\parallel^2+M\mid\mid\mid W\mid\mid\mid^4+M\lambda\label{xyrv1} \end{matrix}$
在区间$[t]$ 上关于$t$ 积分, 可得不等式(3.26). 证毕
(3.28) $\begin{matrix} && \parallel (\nabla\bar{z}^\lambda_t,\lambda{\rm div} \bar{E}^\lambda_t,\lambda^2\nabla{\rm div} \bar{E}^\lambda_t,\nabla \bar{v}^\lambda_t,\nabla\bar{\rho}^\lambda_t)\parallel^2\nonumber \\ &&+k\int_0^t\parallel (\Delta \bar{z}^\lambda_t,{\rm div}\bar{E}^\lambda_t,\lambda \nabla{\rm div}\bar{E}^\lambda_t,\lambda^2 \Delta{\rm div}\bar{E}^\lambda_t,\Delta \bar{v}^\lambda_t,\nabla{\rm div}\bar{v}^\lambda_t) \parallel^2{\rm d}t\nonumber \\ &\leq& \parallel (\nabla\bar{z}^\lambda_t,\lambda{\rm div} \bar{E}^\lambda_t,\lambda^2\nabla{\rm div} \bar{E}^\lambda_t,\nabla \bar{v}^\lambda_t,\nabla\bar{\rho}^\lambda_t)\parallel^2(t=0)\nonumber \\ && +M\int_0^t\parallel(\bar{E}^\lambda,\bar{E}^\lambda_{t}, {\rm div}\bar{E}^\lambda,\lambda\nabla{\rm div}\bar{E}^\lambda,\bar{v}^\lambda,\nabla \bar{v}^\lambda,\bar{v}^\lambda_{t},\nabla \bar{v}^\lambda_{t},{\rm div}\bar{v}^\lambda,\nabla{\rm div}\bar{v}^\lambda)\parallel^2{\rm d}t\nonumber \\ && +M\int_{T^3}(\bar{z}^\lambda,\bar{z}^\lambda_{t},\nabla \bar{z}^\lambda,\nabla \bar{z}^\lambda_{t},\bar{\rho}^\lambda,\bar{\rho}^\lambda_t,\nabla\bar{\rho}^\lambda,\nabla\bar{\rho}^\lambda_t,\triangle\bar{\rho}^\lambda){\rm d}x\nonumber \\ && +M\int _0^t(\mid\mid\mid W\mid\mid\mid^4+\mid\mid\mid W\mid\mid\mid^2G^\lambda(s)){\rm d}s+M\lambda. \label{eq:eq-zevest4} \end{matrix}$
证 对方程(2.12)关于$t$ 求导, 乘以$-\Delta \bar{z}^\lambda_{t}$ , 在区间${{\Bbb T}^3}$ 上关于$x$ 积分, 应用Green 公式, 有
(3.29) $\begin{matrix} &&\frac{1}{2}\frac{\rm d}{{\rm d}t}\parallel \nabla\bar{z}_t^\lambda \parallel^2+\frac{\mu_n+\mu_p}{2}\parallel \triangle \bar{z}_t^\lambda\parallel^2\nonumber\\ &=&-\frac{\mu_n+\mu_p}{2}\int_{{\Bbb T}^3} (A-\nabla \bar{z}^\lambda)_t\triangle \bar{z}_t^\lambda {\rm d}x-\frac{\mu_n-\mu_p}{2}\int_{{\Bbb T}^3} B_t\triangle \bar{z}^\lambda_{t}{\rm d}x\nonumber \\ &&+\int_{{\Bbb T}^3} (z\bar{v}^\lambda+\bar{z}^\lambda v)_t\triangle \bar{z}^\lambda_t{\rm d}x-\frac{\mu_n-\mu_p}{2}\int_{{\Bbb T}^3}(\bar{z}^\lambda \bar{E}^\lambda)_t\triangle \bar{z}^\lambda_t{\rm d}x\nonumber \\ &&+\lambda^2\frac{\mu_n+\mu_p}{2}\int_{{\Bbb T}^3}(\bar{E}^\lambda{\rm div}\bar{E}^\lambda)_t\triangle \bar{z}^\lambda_t{\rm d}x+\int_{{\Bbb T}^3}(\bar{z}^\lambda \bar{v}^\lambda)_t\triangle \bar{z}^\lambda_t {\rm d}x. \label{eq:eq-zest25} \end{matrix}$
估计方程(3.29)右端各项. 对于线性项, 应用Cauchy-Schwarz不等式, Sobolev嵌入定理, 见文献[4 ], 线性项可由(3.30)式控制
(3.30) $\begin{matrix} &&\epsilon\parallel( \bar{E}^\lambda,{\rm div}\bar{E}^\lambda,\nabla{\rm div}\bar{E}^\lambda,\bar{E}^\lambda_{t},{\rm div}\bar{E}^\lambda_{t},\bar{z}^\lambda,\nabla \bar{z}^\lambda,\nabla \bar{z}^\lambda_{t},\bar{z}^\lambda_{t},\bar{v}^\lambda,\bar{v}^\lambda_{t},{\rm div}\bar{v}^\lambda,\nabla\bar{v}^\lambda_t) \parallel^2\nonumber \\ && +\epsilon\frac{\mu_n+\mu_p}{2}\parallel\Delta \bar{z}^\lambda_{t}\parallel^2+ M(\epsilon)\lambda^4\parallel (\nabla{\rm div}\bar{E}^\lambda_{t},\Delta{\rm div}\bar{E}^\lambda_{t}) \parallel^2+M\lambda.\label{eq:eq-zest26} \end{matrix}$
(3.31) $ \begin{matrix} && \int_{{\Bbb T}^3} {\rm div}{(\bar{z}^\lambda\bar{v}^\lambda)_t}\Delta \bar{z}^\lambda_{t} {\rm d}x\nonumber \\ &=&\int_{{\Bbb T}^3} \nabla \bar{z}^\lambda\bar{v}^\lambda_{t}\Delta \bar{z}^\lambda_{t} {\rm d}x+\int_{{\Bbb T}^3} \nabla \bar{z}^\lambda_{t}\bar{v}^\lambda\Delta \bar{z}^\lambda_{t} {\rm d}x+\int_{{\Bbb T}^3} \bar{z}^\lambda_{t}{\rm div}\bar{v}^\lambda\Delta \bar{z}^\lambda_{t} {\rm d}x+\int_{{\Bbb T}^3} \bar{z}^\lambda{\rm div}\bar{v}^\lambda_{t}\Delta \bar{z}^\lambda_{t} {\rm d}x\nonumber\\ &\leq &M(\epsilon)\parallel \nabla \bar{z}^\lambda\bar{v}^\lambda_{t} \parallel^2+ M(\epsilon)\parallel \nabla \bar{z}^\lambda_{t}\bar{v}^\lambda \parallel^2+ M(\epsilon)\parallel \bar{z}^\lambda_{t}{\rm div}\bar{v}^\lambda \parallel^2\nonumber \\ && + M(\epsilon)\parallel \bar{z}^\lambda{\rm div}\bar{v}^\lambda_{t} \parallel^2+\epsilon\parallel \Delta \bar{z}^\lambda_{t} \parallel^2\nonumber\\ &\leq & M(\epsilon)M_s\parallel \nabla \bar{z}^\lambda\parallel_{H^1}^2 \parallel \bar{v}^\lambda_{t} \parallel_{H^1}^2+M(\epsilon)M_s\parallel \nabla \bar{z}^\lambda_{t}\parallel^2 \parallel \bar{v}^\lambda \parallel_{H^2}^2\nonumber \\ && +M(\epsilon)M_s\parallel \bar{z}^\lambda_{t}\parallel^2_{H^2} \parallel {\rm div}\bar{v}^\lambda \parallel_{H^1}^2+M(\epsilon)M_s\parallel \bar{z}^\lambda\parallel^2_{H^2} \parallel{\rm div} \bar{v}^\lambda_t \parallel^2 +\epsilon\parallel \Delta \bar{v}^\lambda_{t} \parallel^2\nonumber \\ &\leq&\epsilon\parallel \Delta \bar{z}^\lambda_{t} \parallel^2+M\mid\mid\mid W\mid\mid\mid^4,\label{eq:eq-zest27} \end{matrix}$
(3.32) $ \begin{matrix} &&-\frac{\mu_n-\mu_p}{2}\int_{{\Bbb T}^3} {\rm div}{(\bar{z}^\lambda\bar{E}^\lambda)_t}\Delta \bar{z}^\lambda_{t} {\rm d}x\nonumber\\ & =&-\frac{\mu_n-\mu_p}{2}\int_{{\Bbb T}^3} ({\nabla \bar{z}^\lambda_{t}\bar{E}^\lambda+\bar{z}^\lambda_{t}{\rm div} \bar{E}^\lambda+\nabla \bar{z}^\lambda\bar{E}^\lambda_{t}+\bar{z}^\lambda{\rm div}\bar{E}^\lambda_{t}})\Delta \bar{E}^\lambda_{t} {\rm d}x\nonumber\\ &\leq & M(\epsilon)(\parallel \nabla \bar{z}^\lambda_{t}\bar{E}^\lambda \parallel^2+\parallel \bar{z}^\lambda_{t}{\rm div} \bar{E}^\lambda \parallel^2)\nonumber \\ && +M(\epsilon)(\parallel \nabla \bar{z}^\lambda\bar{E}^\lambda_{t} \parallel^2+\parallel \bar{z}^\lambda{\rm div}\bar{E}^\lambda_{t} \parallel^2)+\epsilon\frac{\mu_n-\mu_p}{2}\parallel \Delta \bar{z}^\lambda_{t} \parallel^2\nonumber \\ & \leq& M(\epsilon)\parallel \nabla \bar{z}^\lambda_{t}\parallel^2 \parallel \bar{E}^\lambda \parallel_{H^2}^2+M(\epsilon)\parallel \bar{z}^\lambda_{t}\parallel_{H^1}^2 \parallel{\rm div} \bar{E}^\lambda \parallel_{H^1}^2\nonumber\\ && +M(\epsilon)\parallel \nabla \bar{z}^\lambda\parallel_{H^1}^2 \parallel \bar{E}^\lambda_{t} \parallel_{H^1}^2 +M(\epsilon)\parallel \bar{z}^\lambda\parallel_{H^2}^2\parallel{\rm div}\bar{E}^\lambda_{t} \parallel^2+\epsilon\frac{\mu_n-\mu_p}{2}\parallel \Delta \bar{z}^\lambda_{t} \parallel^2\nonumber \\ & \leq&\epsilon\frac{\mu_n-\mu_p}{2}\parallel \Delta \bar{z}^\lambda_{t} \parallel^2+M\mid\mid\mid W\mid\mid\mid^4+M\mid\mid\mid W\mid\mid\mid^2G^\lambda(t),\label{eq:eq-zest28} \end{matrix}$
(3.33) $ \begin{matrix} && \lambda^2\frac{\mu_n+\mu_p}{2}\int_{{\Bbb T}^3}{\rm div}{(\bar{E}^\lambda{\rm div}\bar{E}^\lambda)}_t\Delta \bar{z}^\lambda_{t}{\rm d}x \nonumber \\ & =&\lambda^2\frac{\mu_n+\mu_p}{2}\int_{{\Bbb T}^3}({\rm div}{\bar{E}^\lambda_{t}{\rm div}\bar{E}^\lambda}+{\bar{E}^\lambda_{t}\nabla{\rm div}\bar{E}^\lambda}+{{\rm div}\bar{E}^\lambda{\rm div}\bar{E}^\lambda_{t}}+{\bar{E}^\lambda\nabla{\rm div}\bar{E}^\lambda_{t}}) \Delta \bar{z}^\lambda_{t} {\rm d}x \nonumber \\ &\leq& \lambda^4M(\epsilon)( \parallel {\rm div}{\bar{E}^\lambda_{t}{\rm div}\bar{E}^\lambda}\parallel^2+\parallel{\bar{E}^\lambda_{t}\nabla{\rm div}\bar{E}^\lambda}\parallel^2)\nonumber \\ &&+\lambda^4M(\epsilon)(\parallel{{\rm div}\bar{E}^\lambda{\rm div}\bar{E}^\lambda_{t}}\parallel^2+\parallel{\bar{E}^\lambda\nabla{\rm div}\bar{E}^\lambda_{t}} \parallel^2)+\epsilon\frac{\mu_n+\mu_p}{2}\parallel \Delta \bar{z}^\lambda_{t} \parallel^2\nonumber \\ &\leq& \lambda^4M(\epsilon)( \parallel {\rm div}{\bar{E}^\lambda_{t}\parallel_{H^1}^2\parallel{\rm div}\bar{E}^\lambda}\parallel_{H^1}^2+\parallel{\bar{E}^\lambda_{t}\parallel_{H^1}^2\parallel\nabla{\rm div}\bar{E}^\lambda}\parallel_{H^1}^2)\nonumber \\ && +\lambda^4M(\epsilon)(\parallel{{\rm div}\bar{E}^\lambda\parallel_{H^1}^2\parallel{\rm div}\bar{E}^\lambda_{t}}\parallel_{H^1}^2+\parallel{\bar{E}^\lambda\parallel_{H^2}^2\parallel\nabla{\rm div}\bar{E}^\lambda_{t}} \parallel^2)+\epsilon\frac{\mu_n+\mu_p}{2}\parallel \Delta \bar{z}^\lambda_{t} \parallel^2\nonumber \\ & \leq&\epsilon\frac{\mu_n+\mu_p}{2}\parallel \Delta \bar{z}^\lambda_{t} \parallel^2+M\mid\mid\mid W\mid\mid\mid^4+M\mid\mid\mid W\mid\mid\mid^2G^\lambda(t)+M\lambda^2\parallel \bar{E}^\lambda_{t}\parallel^2_{H^2}\mid\mid\mid W\mid\mid\mid^2.~~~~~~~\label{eq:eq-zest29} \end{matrix}$
(3.34) $\begin{matrix} &&\frac{\rm d}{{\rm d}t}\parallel\nabla \bar{z}^\lambda_{t} \parallel^2+k\parallel \Delta \bar{z}^\lambda_{t}\parallel^2\nonumber \\ &\leq& M\parallel(\bar{z}^\lambda,\bar{z}^\lambda_{t},\nabla \bar{z}^\lambda,\nabla \bar{z}^\lambda_{t},\bar{E}^\lambda,\bar{E}^\lambda_{t},{\rm div}\bar{E}^\lambda,{\rm div}\bar{E}^\lambda_{t},\nabla{\rm div}\bar{E}^\lambda,\bar{v}^\lambda,\bar{v}^\lambda_{t},{\rm div}\bar{v}^\lambda,\nabla\bar{v}^\lambda_{t})\parallel^2\nonumber \\ && +M\lambda^4\parallel(\nabla {\rm div}\bar{E}^\lambda_{t},\Delta{\rm div}\bar{E}^\lambda_{t})\parallel^2+M\mid\mid\mid W\mid\mid\mid^4+M\lambda^2\parallel E_{R,t}\parallel^2_{H^2}\mid\mid\mid W\mid\mid\mid^2\nonumber \\ &&+M\mid\mid\mid W\mid\mid\mid^2G^\lambda(t)+M\lambda.\label{eq:eq-zest32} \end{matrix}$
对方程(2.13)两端关于$t$ 求导, 乘以$-\Delta\bar{\phi}^\lambda_{t}={\rm div}\bar{E}^\lambda_{t}$ , 在区间${{\Bbb T}^3}$ 上关于$x$ 积分, 有
(3.35) $\begin{matrix} &&\frac{\lambda^2}{2}\frac{\rm d}{{\rm d}t}\parallel {\rm div}\bar{E}^\lambda_{t} \parallel^2+\lambda^2\parallel \nabla{\rm div} \bar{E}^\lambda_{t}\parallel^2+\frac{\mu_n+\mu_p}{2}\int_{{\Bbb T}^3} z|{\rm div}\bar{E}^\lambda_{t}|^2{\rm d}x\nonumber \\ &=&\frac{\mu_n+\mu_p}{2}\int_{{\Bbb T}^3}B_t\nabla{\rm div}\bar{E}^\lambda_{t}{\rm d}x+\frac{\mu_n+\mu_p}{2}\int_{{\Bbb T}^3} z |{\rm div}\bar{E}^\lambda_{t}|^2{\rm d}x-\frac{\mu_n+\mu_p}{2}\int_{{\Bbb T}^3} z_t\bar{E}^\lambda\nabla{\rm div} \bar{E}^\lambda_{t}{\rm d}x\nonumber \\ &&-\frac{\mu_n-\mu_p}{2}\int_{{\Bbb T}^3}A_t\nabla{\rm div}\bar{E}^\lambda_{t}{\rm d}x+\int_{{\Bbb T}^3}(v({\rm div}\bar{E}^\lambda+{\rm div}\varepsilon))_t\nabla{\rm div}\bar{E}^\lambda_{t}{\rm d}x\nonumber\\ &&-\int_{{\Bbb T}^3} ((\bar{v}^\lambda+v){\rm div}\varepsilon)_t \nabla{\rm div}\bar{E}^\lambda_{t}{\rm d}x +\int_{{\Bbb T}^3} D\bar{v}^\lambda_t \nabla{\rm div}\bar{E}^\lambda_{t}{\rm d}x-\frac{\mu_n+\mu_p}{2}\int_{{\Bbb T}^3} {(\bar{z}^\lambda\bar{E}^\lambda)_t}\nabla{\rm div}\bar{E}^\lambda_{t} {\rm d}x\nonumber\\ && -\int_{{\Bbb T}^3} (\bar{v}^\lambda{\rm div}\bar{E}^\lambda)_t\nabla{\rm div}\bar{E}^\lambda_{t} {\rm d}x +\lambda^2\frac{\mu_n-\mu_p}{2}\int_{{\Bbb T}^3} {(\bar{E}^\lambda{\rm div}\bar{E}^\lambda)_t}\nabla{\rm div}\bar{E}^\lambda_{t} {\rm d}x.\label{eq:eq-eest20} \end{matrix}$
估计方程(3.35)右端各项. 对线性项应用Cauchy-Schwarz不等式, Green公式以及Sobolev嵌入不等式, $\bar{E}^\lambda_t=-\nabla\bar{\phi}^\lambda_t$ , 线性项可由(3.36)式控制
(3.36) $\begin{matrix} &&\epsilon\parallel(\bar{z}^\lambda, \bar{z}^\lambda_t,\nabla \bar{z}^\lambda,\nabla \bar{z}^\lambda_t, \bar{E}^\lambda,{\rm div}\bar{E}^\lambda,\bar{E}^\lambda_{t},{\rm div}\bar{E}^\lambda_{t},\bar{v}^\lambda,\bar{v}^\lambda_{t},{\rm div}\bar{v}^\lambda,\nabla\bar{v}^\lambda_{t}) \parallel^2\nonumber \\ &&+M(\epsilon)\lambda^4 \parallel (\nabla{\rm div}\bar{E}^\lambda,\nabla{\rm div}\bar{E}^\lambda_{t}) \parallel^2+M\lambda.\label{eq:eq-eest21} \end{matrix}$
对非线性项, 应用Cauchy-Schwarz不等式和Green公式, 有
(3.37) $\begin{matrix} &&-\frac{\mu_n+\mu_p}{2}\int_{{\Bbb T}^3} {\rm div}{(\bar{z}^\lambda\bar{E}^\lambda)_t}{\rm div}\bar{E}^\lambda_{t} {\rm d}x\nonumber \\ & =&-\frac{\mu_n+\mu_p}{2}\int_{{\Bbb T}^3} {(\bar{z}^\lambda_{t}{\rm div}\bar{E}^\lambda+\nabla \bar{z}^\lambda\bar{E}^\lambda_{t}+\nabla \bar{z}^\lambda_{t}\bar{E}^\lambda+\bar{z}^\lambda{\rm div}\bar{E}^\lambda_{t})}{\rm div}\bar{E}^\lambda_{t} {\rm d}x\nonumber\\ & \leq& \epsilon\parallel {\rm div}\bar{E}^\lambda_{t} \parallel^2+M(\epsilon)(\|\bar{z}^\lambda_{t}{\rm div}\bar{E}^\lambda\|^2+\|\nabla \bar{z}^\lambda\bar{E}^\lambda_{t}\|^2+\|\nabla \bar{z}^\lambda_{t}\bar{E}^\lambda\|^2+\|\bar{z}^\lambda{\rm div}\bar{E}^\lambda_{t}\|^2)\nonumber \\ & \leq& \epsilon\parallel {\rm div}\bar{E}^\lambda_{t} \parallel^2+M(\epsilon)M_s\|\bar{z}^\lambda_{t}\|_{H^2}^2\|{\rm div}\bar{E}^\lambda\|^2+M(\epsilon)M_s\|\nabla \bar{z}^\lambda\|_{H^1}^2\|\bar{E}^\lambda_{t}\|_{H^1}^2\nonumber \\ &&+M(\epsilon)M_s\|\nabla \bar{z}^\lambda_{t}\|_{H^1}^2\|\bar{E}^\lambda\|_{H^1}^2+M(\epsilon)M_s\|\bar{z}^\lambda\|_{H^2}^2\|{\rm div}\bar{E}^\lambda_{t}\|^2\nonumber \\ &\leq &\epsilon\parallel {\rm div}\bar{E}^\lambda_{t} \parallel^2+M\mid\mid\mid W\mid\mid\mid^2G^\lambda(t)+M \mid\mid\mid W\mid\mid\mid^4,\label{eq:eq-eest22} \end{matrix}$
(3.38) $\begin{matrix} && \lambda^2\frac{\mu_n-\mu_p}{2}\int_{{\Bbb T}^3}{{\rm div}(\bar{E}^\lambda{\rm div}\bar{E}^\lambda)_t}{\rm div}\bar{E}^\lambda_{t} {\rm d}x\nonumber \\ & =&\lambda^2\frac{\mu_n-\mu_p}{2}\int_{{\Bbb T}^3} (\nabla{\rm div}\bar{E}^\lambda \bar{E}^\lambda_{t}+\frac 32{\rm div}\bar{E}^\lambda{\rm div}\bar{E}^\lambda_{t}){\rm div}\bar{E}^\lambda_{t}{\rm d}x\nonumber \\ & \le&\epsilon \|{\rm div}\bar{E}^\lambda_{t}\|^2+M(\epsilon)\lambda^4(\|\nabla{\rm div}\bar{E}^\lambda \bar{E}^\lambda_{t}\|^2+\|{\rm div}\bar{E}^\lambda{\rm div}\bar{E}^\lambda_{t}\|^2)\nonumber \\ & \le&\epsilon \|{\rm div}\bar{E}^\lambda_{t}\|^2+M(\epsilon)M_s\lambda^4(\|\nabla{\rm div}\bar{E}^\lambda\|^2\| \bar{E}^\lambda_{t}\|_{H^2}^2+\|{\rm div}\bar{E}^\lambda\|_{H^1}\|{\rm div}\bar{E}^\lambda_{t}\|_{H^1}^2)\nonumber\\ & \leq &\epsilon\parallel {\rm div}\bar{E}^\lambda_{t} \parallel^2+M\lambda^2\parallel \bar{E}^\lambda_{t}\parallel^2_{H^2}\mid\mid\mid W\mid\mid\mid^2+M\mid\mid\mid W\mid\mid\mid^2G^\lambda(t),\label{eq:eq-eest23} \end{matrix}$
(3.39) $\begin{matrix} && -\lambda^2 \int_{{\Bbb T}^3} {\rm div}(\bar{v}^\lambda{\rm div}\bar{E}^\lambda )_t{\rm div}\bar{E}^\lambda_{t} {\rm d}x\nonumber \\ & =&-\lambda^2 \int_{{\Bbb T}^3}( \bar{v}^\lambda_{t}\nabla{\rm div}\bar{E}^\lambda+\bar{v}^\lambda\nabla{\rm div}\bar{E}^\lambda_{t}+{\rm div}\bar{v}^\lambda_{t}{\rm div}\bar{E}^\lambda+{\rm div}\bar{v}^\lambda{\rm div}\bar{E}^\lambda_t){\rm div}\bar{E}^\lambda_{t} {\rm d}x\nonumber \\ &\leq& M(\epsilon)\lambda^4\parallel \bar{v}^\lambda_{t} \nabla {\rm div}\bar{E}^\lambda\parallel^2+M(\epsilon)\lambda^4\parallel \bar{v}^\lambda \nabla {\rm div}\bar{E}^\lambda_t\parallel^2\nonumber \\ && +M(\epsilon)\lambda^4\parallel {\rm div}\bar{v}^\lambda_{t} {\rm div}\bar{E}^\lambda\parallel^2+M(\epsilon)\lambda^4\parallel {\rm div}\bar{v}^\lambda {\rm div}\bar{E}^\lambda_t\parallel^2+\epsilon\parallel{\rm div}E_{R,t}\parallel^2\nonumber \\ &\leq & M(\epsilon)\lambda^4\parallel \bar{v}^\lambda_{t}\parallel_{H^2}^2 \parallel \nabla{\rm div}\bar{E}^\lambda\parallel^2+M(\epsilon)\lambda^4\parallel \bar{v}^\lambda\parallel_{H^2}^2 \parallel \nabla{\rm div}\bar{E}^\lambda+t\parallel^2\nonumber \\ &&M(\epsilon)\lambda^4\parallel {\rm div}\bar{v}^\lambda\parallel_{H^1}^2 \parallel {\rm div}\bar{E}^\lambda\parallel^2_{H^1}+M(\epsilon)\lambda^4\parallel {\rm div}\bar{v}^\lambda\parallel_{H^1}^2 \parallel {\rm div}\bar{E}^\lambda\parallel^2_{H^1}+\epsilon\parallel {\rm div}\bar{E}^\lambda_{t} \parallel^2\nonumber \\ & \leq& \epsilon\parallel {\rm div}\bar{E}^\lambda_{t} \parallel^2+M\mid\mid\mid W\mid\mid\mid^4+M\mid\mid\mid W\mid\mid\mid^2G^\lambda(t).\label{eq:eq-eest24} \end{matrix}$
联合不等式(3.35)-(3.39), 并假设$\lambda$ 足够小, 可得
(3.40) $\begin{matrix} &&\lambda^2\frac{\rm d}{{\rm d}t}\parallel {\rm div}\bar{E}^\lambda_{t} \parallel^2+k\parallel ({\rm div}\bar{E}^\lambda_{t},\lambda \nabla{\rm div} \bar{E}^\lambda_{t})\parallel^2\nonumber \\ & \leq & M(\epsilon)\parallel(\bar{z}^\lambda,\bar{z}^\lambda_{t},\nabla\bar{z}^\lambda,\nabla \bar{z}^\lambda_{t},\bar{E}^\lambda,\bar{v}^\lambda,\bar{v}^\lambda_{t},{\rm div}\bar{v}^\lambda,{\rm div}\bar{v}^\lambda_{t})\parallel^2+M\lambda^2\parallel(\bar{E}^\lambda,\bar{E}^\lambda_{t},{\rm div}\bar{E}^\lambda)\parallel^2\nonumber \\ &&+M\mid\mid\mid W\mid\mid\mid^4+M\mid\mid\mid W\mid\mid\mid^2G^\lambda(t)+M\lambda.\label{eq:eq-eest26} \end{matrix}$
对方程(2.13)关于$t$ 求导, 乘以$\lambda^2\Delta{\rm div}\bar{E}^\lambda_{t}$ , 在区间${{\Bbb T}^3}$ 上关于$x$ 积分, 有
(3.41) $\begin{matrix} &&\frac{\lambda^4}{2}\frac{\rm d}{{\rm d}t}\parallel\nabla {\rm div}\bar{E}^\lambda_{t} \parallel^2+\lambda^4\parallel \triangle{\rm div} \bar{E}^\lambda_{t}\parallel^2+\frac{\mu_n+\mu_p}{2}\lambda^2\int_{{\Bbb T}^3} z|\nabla{\rm div}\bar{E}^\lambda_{t}|^2{\rm d}x\nonumber \\ & =&\frac{\mu_n+\mu_p}{2}\int_{{\Bbb T}^3}{\rm div}B_t\triangle{\rm div}\bar{E}^\lambda_{t}{\rm d}x+\frac{\mu_n+\mu_p}{2}\lambda^2\int_{{\Bbb T}^3} z |\nabla{\rm div}\bar{E}^\lambda_{t}|^2{\rm d}x\nonumber\\ && -\frac{\mu_n+\mu_p}{2}\int_{{\Bbb T}^3} z_t\bar{E}^\lambda\triangle{\rm div} \bar{E}^\lambda_{t}{\rm d}x-\frac{\mu_n-\mu_p}{2}\int_{{\Bbb T}^3}{\rm div}A_t\triangle{\rm div}\bar{E}^\lambda_{t}{\rm d}x\nonumber\\ && +\lambda^2\int_{{\Bbb T}^3}(v({\rm div}\bar{E}^\lambda+{\rm div}\varepsilon))_t\triangle{\rm div}\bar{E}^\lambda_{t}{\rm d}x-\lambda^2\int_{{\Bbb T}^3} ((\bar{v}^\lambda+v){\rm div}\varepsilon)_t \triangle{\rm div}\bar{E}^\lambda_{t}{\rm d}x\nonumber \\ &&+\int_{{\Bbb T}^3} D\bar{v}^\lambda_t \triangle{\rm div}\bar{E}^\lambda_{t}{\rm d}x-\frac{\mu_n+\mu_p}{2}\int_{T^3} {(\bar{z}^\lambda\bar{E}^\lambda)_t}\triangle{\rm div}\bar{E}^\lambda_{t} {\rm d}x\nonumber\\ &&-\lambda^2\int_{{\Bbb T}^3} (\bar{v}^\lambda{\rm div}\bar{E}^\lambda)_t\triangle{\rm div}\bar{E}^\lambda_{t} {\rm d}x +\lambda^2\frac{\mu_n-\mu_p}{2}\int_{{\Bbb T}^3} {(\bar{E}^\lambda{\rm div}\bar{E}^\lambda)_t}\triangle{\rm div}\bar{E}^\lambda_{t} {\rm d}x.\label{eq:eq-ee15} \end{matrix}$
由于方程(3.41)与方程(3.35)完全类似, 同理可得
(3.42) $\begin{matrix} &&\lambda^4 \frac{\rm d}{{\rm d}t}\parallel\nabla{\rm div} \bar{E}^\lambda_{t} \parallel^2+k\lambda^2\parallel( \nabla{\rm div} \bar{E}^\lambda_{t},\lambda\Delta {\rm div} \bar{E}^\lambda_{t}) \parallel^2\nonumber \\ & \leq & M\parallel(\bar{z}^\lambda,\bar{z}^\lambda_{t},\nabla \bar{z}^\lambda,\nabla \bar{z}^\lambda_{t}, \bar{v}^\lambda,\bar{v}^\lambda_{t})\parallel^2+ M\lambda^4\parallel(\nabla{\rm div}\bar{E}^\lambda,{\rm div}\bar{E}^\lambda_{t})\parallel^2\nonumber \\ && +M\lambda^2\parallel(\bar{E}^\lambda,\bar{E}^\lambda_{t},{\rm div}\bar{E}^\lambda)\parallel^2+\epsilon\frac{\mu_n+\mu_p}{2}\parallel\Delta{\rm div}\bar{E}^\lambda_{t}\parallel^2\nonumber \\ && +M\lambda^2\parallel \bar{E}^\lambda_{t}\parallel^2_{H^2}\mid\mid\mid W\mid\mid\mid^2+ M\mid\mid\mid W\mid\mid\mid^4+M\mid\mid\mid \mid W\mid\mid\mid^2G^\lambda(t)+M\lambda.\label{eq:eq-ee217} \end{matrix}$
对方程(2.14)关于$t$ 求导, 乘以$-\Delta \bar{v}^\lambda_{t}$ , 在区间${{\Bbb T}^3}$ 上关于$x$ 积分, 应用分部积分, 有
(3.43) $\begin{matrix} && \int_{{\Bbb T}^3}(\rho+\bar{\rho}^\lambda)\bar{v}^\lambda_{tt}\triangle\bar{v}^\lambda_t {\rm d}x+A\frac{\rm d}{{\rm d}t}\parallel\nabla\bar{\rho}^\lambda_t\parallel^2+\mu\parallel\triangle\bar{v}^\lambda_t\parallel^2+\parallel\nabla{\rm div}\bar{v}^\lambda_t\parallel^2\nonumber \\ & =&-\int_{{\Bbb T}^3}(\bar{\rho}^\lambda v_t+\bar{\rho}^\lambda v \nabla v+\rho\bar{v}^\lambda \nabla v+\rho v \nabla\bar{v}^\lambda+\rho_t\bar{v}^\lambda_t )_t\triangle\bar{v}^\lambda_t{\rm d}x-\int_{{\Bbb T}^3}D\bar{E}^\lambda_t\triangle \bar{v}^\lambda_t {\rm d}x\nonumber \\ && -\lambda^2\int_{{\Bbb T}^3}(\bar{E}^\lambda{\rm div}\varepsilon+\varepsilon{\rm div}\bar{E}^\lambda+\varepsilon{\rm div}\varepsilon)_t\triangle\bar{v}_t^\lambda {\rm d}x-2A\int_{{\Bbb T}^3}(\bar{\rho}^\lambda{\rm div}v+v \nabla\bar{\rho}^\lambda)_t\nabla\rho^\lambda_t{\rm d}x\nonumber \\ && -2A\int_{{\Bbb T}^3}(\rho_t{\rm div}\bar{v}^\lambda+\bar{v}^\lambda \nabla\rho_t)\nabla\bar{\rho}^\lambda_t{\rm d}x-2A\int_{{\Bbb T}^3}(\rho \bar{v}^\lambda \nabla\bar{v}^\lambda+\bar{\rho}^\lambda v \nabla\bar{v}^\lambda+\bar{\rho}^\lambda\bar{v}^\lambda\nabla\bar{v}^\lambda)_t \triangle\bar{v}_t^\lambda {\rm d}x\nonumber \\ &&-\int_{{\Bbb T}^3}((\bar{\rho}^\lambda\bar{v}^\lambda\nabla v)_t+\bar{\rho}_t^\lambda\bar{v}^\lambda_t+\lambda^2(\bar{E}^\lambda{\rm div}\bar{E}^\lambda)_t)\triangle\bar{v}^\lambda_t {\rm d}x-2A\int_{{\Bbb T}^3}({\rm div}\bar{v}^\lambda\bar{\rho}^\lambda_t+\bar{v}^\lambda\nabla\bar{\rho}^\lambda_t)\nabla\bar{\rho}^\lambda_t{\rm d}x.\label{eq:eq-vest16} \end{matrix}$
估计方程(3.43)右端各项. 对线性项应用Cauchy-Schwarz不等式以及Sobolev嵌入定理, 线性项可由(3.44)式控制
(3.44) $\begin{matrix} M \parallel &&(\bar{v}^\lambda,\nabla \bar{v}^\lambda,\bar{v}^\lambda_{t},\nabla \bar{v}^\lambda_{t},{\rm div}\bar{v}^\lambda,\nabla{\rm div}\bar{v}^\lambda, \Delta \bar{v}^\lambda_{t},\nabla{\rm div}\bar{v}^\lambda_t,\bar{\rho}^\lambda,\bar{\rho}^\lambda_t,\nabla\bar{\rho}^\lambda,\nabla\bar{\rho}^\lambda_t,\triangle\bar{\rho}^\lambda)\parallel^2\nonumber \\ && +M\lambda^2\parallel( \bar{E}^\lambda,{\rm div}\bar{E}^\lambda)\parallel^2+M(\epsilon)\lambda^4\parallel(\bar{E}^\lambda_{t},{\rm div}\bar{E}^\lambda_{t}) \parallel^2+M\lambda.\label{eq:eq-vest17} \end{matrix}$
对非线性项, 应用方程(2.15)的导数, Cauchy-Schwarz不等式Sobolev嵌入定理, 有
(3.45) $\begin{matrix} &&-2A\int_{{\Bbb T}^3}(\bar{\rho}^\lambda{\rm div}v+v \nabla\bar{\rho}^\lambda)_t\nabla\rho^\lambda_t{\rm d}x-2A\int_{{\Bbb T}^3}(\rho_t{\rm div}\bar{v}^\lambda+\bar{v}^\lambda \nabla\rho_t)\nabla\bar{\rho}^\lambda_t{\rm d}x\nonumber \\ && -2A\int_{{\Bbb T}^3}(\rho \bar{v}^\lambda \nabla\bar{v}^\lambda+\bar{\rho}^\lambda v \nabla\bar{v}^\lambda +\bar{\rho}^\lambda\bar{v}^\lambda\nabla\bar{v}^\lambda)_t\triangle \bar{v}_t^\lambda {\rm d}x -\int_{{\Bbb T}^3}((\bar{\rho}^\lambda\bar{v}^\lambda\nabla v)_t+\bar{\rho}_t^\lambda\bar{v}^\lambda_t)\triangle\bar{v}^\lambda_t {\rm d}x\nonumber \\ && -\lambda^2\int_{{\Bbb T}^3}(\bar{E}^\lambda{\rm div}\bar{E}^\lambda)_t)\triangle\bar{v}^\lambda_t {\rm d}x -2A\int_{{\Bbb T}^3}({\rm div}\bar{v}^\lambda\bar{\rho}^\lambda_t +\bar{v}^\lambda\nabla\bar{\rho}^\lambda_t)\nabla\bar{\rho}^\lambda_t{\rm d}x\nonumber \\ &\leq& \epsilon \parallel(\triangle\bar{v}^\lambda_t,\nabla{\rm div}\bar{v}^\lambda,\nabla{\rm div}\bar{v}^\lambda_t,\nabla\bar{\rho}^\lambda,\nabla\bar{\rho}^\lambda_t)\parallel^2+M\parallel\bar{v}^\lambda\nabla\bar{v}^\lambda_t \parallel^2+M\parallel\bar{\rho}^\lambda\bar{v}^\lambda_t\parallel^2\nonumber \\ && +M\parallel\nabla\bar{v}^\lambda\bar{v}^\lambda_t\parallel^2+M\parallel\bar{\rho}^\lambda_t\nabla\bar{v}^\lambda\parallel^2+M\parallel\bar{v}^\lambda\bar{\rho}^\lambda_t\parallel^2\nonumber \\ && +M\parallel\nabla\bar{\rho}^\lambda\nabla\bar{v}^\lambda_t\parallel^2+M\parallel\bar{\rho}^\lambda_t{\rm div}\bar{v}^\lambda\parallel^2+M\parallel\bar{\rho}^\lambda_t\nabla\bar{\rho}^\lambda\parallel^2+M\parallel\bar{\rho}^\lambda_t\bar{v}^\lambda_t\parallel^2\nonumber \\ && +M\parallel\nabla\bar{\rho}^\lambda_t\nabla\bar{v}^\lambda\parallel^2+M\parallel\nabla\bar{\rho}^\lambda{\rm div}\bar{v}^\lambda_t\parallel^2+M\parallel\nabla\bar{\rho}^\lambda{\rm div}\bar{v}^\lambda\parallel^2+M\parallel\nabla\bar{\rho}^\lambda\bar{v}^\lambda_t\parallel^2\nonumber \\ && +M\parallel\bar{\rho}^\lambda\nabla\bar{v}^\lambda_t\parallel^2+\lambda^4M\parallel\bar{E}^\lambda_t{\rm div}\bar{E}^\lambda\parallel^2+M\lambda^4\parallel\bar{E}^\lambda{\rm div}\bar{E}^\lambda_t\parallel^2\nonumber \\ & \leq &\epsilon \parallel(\triangle\bar{v}^\lambda_t,\nabla{\rm div}\bar{v}^\lambda,\nabla{\rm div}\bar{v}^\lambda_t,\nabla\bar{\rho}^\lambda,\nabla\bar{\rho}^\lambda_t)\parallel^2+M\parallel\nabla\bar{v}^\lambda_t \parallel^2\parallel\bar{v}^\lambda\parallel^2_{H^2}\nonumber \\ && +M\parallel\bar{v}^\lambda_t \parallel^2_{H^1}\parallel\bar{\rho}^\lambda\parallel^2_{H^1}+M\parallel\bar{v}^\lambda_t \parallel^2_{H^1}\parallel\nabla\bar{v}^\lambda\parallel^2_{H^1}+M\parallel\bar{\rho}^\lambda_t \parallel^2_{H^1}\parallel\nabla\bar{v}^\lambda\parallel^2_{H^1}\nonumber \\ && +M\parallel\bar{\rho}^\lambda_t \parallel^2\parallel\bar{v}^\lambda\parallel^2_{H^2}+M\parallel\bar{v}^\lambda_t \parallel^2_{H^1}\parallel\bar{\rho}^\lambda_t\parallel^2_{H^1}+M\parallel\bar{\rho}^\lambda_t \parallel^2_{H^1}\parallel{\rm div}\bar{v}^\lambda\parallel^2_{H^1}\nonumber \\ && +M\parallel\bar{\rho}^\lambda_t \parallel^2_{H^1}\parallel\nabla\bar{\rho}^\lambda\parallel^2_{H^1}+M\parallel\nabla\bar{\rho}^\lambda_t \parallel^2_{H^1}\parallel{\rm div}\bar{v}^\lambda\parallel^2_{H^1}+M\parallel\nabla\bar{\rho}^\lambda_t \parallel^2_{H^1}\parallel\nabla\bar{v}^\lambda\parallel^2_{H^1}\nonumber \\ && +M\parallel\nabla\bar{\rho}^\lambda_t \parallel^2_{H^1}\parallel{\rm div}\bar{v}^\lambda_t\parallel^2_{H^1}+M\parallel\nabla\bar{\rho}^\lambda \parallel^2_{H^1}\parallel{\rm div}\bar{v}^\lambda\parallel^2_{H^1}+M\parallel\triangle\bar{\rho}^\lambda\parallel^2\parallel\bar{v}^\lambda_t\parallel^2_{H^2}\nonumber \\ && +M\lambda^4\parallel\bar{E}^\lambda_t \parallel^2_{H^1}\parallel{\rm div}\bar{E}^\lambda\parallel^2_{H^1}+M\lambda^4\parallel\bar{E}^\lambda \parallel^2_{H^2}\parallel{\rm div}\bar{E}^\lambda_t\parallel^2\nonumber \\ & \leq& \epsilon \parallel(\triangle\bar{v}^\lambda_t,\nabla{\rm div}\bar{v}^\lambda,\nabla{\rm div}\bar{v}^\lambda_t,\nabla\bar{\rho}^\lambda,\nabla\bar{\rho}^\lambda_t)\parallel^2+M\mid\mid\mid W\mid\mid\mid^4.\label{eq:eq-vest19} \end{matrix}$
联合不等式(3.43)-(3.45), 并应用不等式(3.46)
(3.46) $\begin{equation} \int_{{\Bbb T}^3}(\rho+\bar{\rho}^\lambda)\bar{v}^\lambda_{tt}\triangle\bar{v}^\lambda_t {\rm d}x\geq \frac{\underline{\rho}}{2}\int_{{\Bbb T}^3}\bar{v}^\lambda_{tt}\triangle\bar{v}^\lambda_t {\rm d}x=\frac{\underline{\rho}}{4}\frac{\rm d}{{\rm d}t}\parallel\nabla\bar{v}^\lambda_t\parallel^2,\label{eq:rvt3} \end{equation}$
$\rho$ 和$\bar{\rho}^\lambda$ 满足下面的不等式
(3.47) $\begin{equation} \frac{\underline{\rho}}{2}\leq \rho-|\bar{\rho}^\lambda|_{L^\infty}\leq\rho+\bar{\rho}^\lambda\leq\overline{\rho}+|\bar{\rho}^\lambda|_{L^\infty}\leq2\overline{\rho},\label{eq:rvt4} \end{equation} $
其中$\underline{\rho}$ 和$\overline{\rho}$ 分别是下确界和上确界, 可得
(3.48) $\begin{matrix} && \epsilon\frac{\rm d}{{\rm d}t}\parallel(\nabla\bar{\rho}^\lambda_t,\nabla\bar{v}^\lambda_t) \parallel^2+k\parallel (\triangle\bar{v}^\lambda_t,\nabla{\rm div}\bar{v}^\lambda_t)\parallel^2 \nonumber \\ & \leq & M\parallel(\bar{v}^\lambda,\bar{v}^\lambda_t,\nabla\bar{v}^\lambda,{\rm div}\bar{v}^\lambda,\nabla\bar{v}^\lambda_t,\nabla{\rm div}\bar{v}^\lambda,\bar{\rho}^\lambda,\bar{\rho}^\lambda_t,\nabla\bar{\rho}^\lambda,\nabla\bar{\rho}^\lambda_t,\triangle\bar{\rho}^\lambda,\bar{E}^\lambda, \bar{E}^\lambda_t)\parallel^2\nonumber \\ && +M\lambda^4\parallel({\rm div}\bar{E}^\lambda,{\rm div} \bar{E}^\lambda_t)\parallel^2 +M\mid\mid\mid W\mid\mid\mid^4+M\lambda.\label{eq:eq-vest20} \end{matrix}$
联合$\delta_4$ (3.32) ,(3.39),(3.42)以及(3.48)式, 在区间$[t]$ 上关于$t$ 积分,取$\delta_4$ $>$ $0$ 足够小, 可得不等式(3.28). 证毕
定理3.6 估计$\parallel(\triangle \bar{z}^\lambda, {\rm div}\bar{E}^\lambda,\lambda\nabla {\rm div}\bar{E}^\lambda, \lambda^2\triangle {\rm div}\bar{E}^\lambda,\triangle \bar{v}^\lambda,\nabla{\rm div}\bar{v}^\lambda,\nabla\bar{\rho}^\lambda)\parallel^2$ , 有
(3.49) $\begin{matrix} &&\parallel(\triangle \bar{z}^\lambda, {\rm div}\bar{E}^\lambda,\lambda\nabla {\rm div}\bar{E}^\lambda, \lambda^2\triangle {\rm div}\bar{E}^\lambda,\triangle \bar{v}^\lambda,\nabla{\rm div}\bar{v}^\lambda,\nabla\bar{\rho}^\lambda)\parallel^2\nonumber \\& \leq& M\parallel(\nabla\bar{z}^\lambda,\nabla\bar{v}^\lambda,\nabla\bar{\rho}^\lambda,\nabla\bar{z}^\lambda_t,\nabla\bar{v}^\lambda_t,\nabla\bar{\rho}^\lambda_t)\parallel^2\nonumber \\ && \lambda\parallel({\rm div}\bar{E}^\lambda,\lambda\nabla{\rm div}\bar{E}^\lambda,{\rm div}\bar{E}^\lambda_t,\lambda\nabla{\rm div}\bar{E}^\lambda_t)\parallel^2+M\mid\mid\mid W\mid\mid\mid^4+M\lambda. \label{eq:lw7} \end{matrix}$
注3.3 不等式(3.49)可由不等式(3.26)和 Green公式而得.
4 证明定理
(4.1) $\begin{equation} \parallel \partial_t^i\bar{z}^\lambda \parallel_{H^2}^2\leq M(\parallel \partial_t^i\bar{z}^\lambda \parallel^2+\parallel \Delta \partial_t^i\bar{z}^\lambda \parallel^2),i=0,1,\label{eq:eq-pf1} \end{equation} $
(4.2) $\begin{equation} \parallel \partial_t^i\bar{v}^\lambda \parallel_{H^2}^2\leq M(\parallel \partial_t^i\bar{v}^\lambda \parallel^2+\parallel \Delta \partial_t^i\bar{v}^\lambda \parallel^2),i=0,1,\label{eq:eq-pf3} \end{equation}$
(4.3) $\begin{equation} \parallel \partial_t^i\bar{\rho}^\lambda \parallel_{H^2}^2\leq M(\parallel \partial_t^i\bar{\rho}^\lambda \parallel^2+\parallel \Delta \partial_t^i\bar{\rho}^\lambda \parallel^2),i=0,1,\label{eq:eq-pf4} \end{equation} $
(4.4) $\begin{equation} \parallel \partial_t^i\bar{E}^\lambda \parallel_{H^2}^2\leq M(\parallel \partial_t^i\bar{E}^\lambda \parallel^2+\parallel \Delta \partial_t^i{\rm div}\bar{T}^\lambda \parallel^2), i=0,1~~s=1,2.\label{eq:eq-pf6} \end{equation} $
由$\Gamma^\lambda(t)$ 和$\mid\mid\mid W\mid\mid\mid$ 的表达式, 并应用不等式(4.1)-(4.4), 知存在独立于$\lambda$ 的两个正常数$C_1$ 和$C_2$ , 使得
(4.5) $\begin{equation} C_1\mid\mid\mid W\mid\mid\mid^2\leq\Gamma^\lambda(t)\leq C_2\mid\mid\mid W\mid\mid\mid^2.\label{eq:equiv-1} \end{equation} $
应用不等式(4.5), 并计算(3.2)+$\delta$ (3.4) +(3.25)+$\delta_1$ (3.26) +$\delta_2$ (3.28) +(3.49), 取$\delta,\delta_1,\delta_2,\lambda$ 足够小, 可得
(4.6) $\begin{matrix} \Gamma^\lambda(t)+\int_0^tG^\lambda(s){\rm d}s &\leq & M\tilde{\Gamma}^\lambda(t=0)+M\int_0^t(\Gamma^\lambda(s)+(\Gamma^\lambda(s))^2){\rm d}s\nonumber\\ && +M(\Gamma^\lambda(t))^2 +M\int_0^t\Gamma^\lambda(s)G^\lambda(s){\rm d}s+M\lambda,\label{eq:txg3-1} \end{matrix}$
(4.7) $\begin{matrix} \tilde{\Gamma^\lambda}(0)&=&\parallel(\bar{z}^\lambda,\bar{v}^\lambda,\bar{\rho}^\lambda,\bar{z}_t^\lambda,\bar{v}_t^\lambda,\bar{\rho}_t^\lambda,\nabla\bar{z}^\lambda,\nabla\bar{v}^\lambda,\nabla\bar{\rho}^\lambda,\nabla\bar{z}_t^\lambda,\nabla\bar{v}_t^\lambda,\nabla\bar{\rho}_t^\lambda)\parallel^2(t=0)\nonumber \\ &&+\lambda^2\parallel (\bar{E}^\lambda,\lambda{\rm div}\bar{E}^\lambda,\bar{E}_t^\lambda,\lambda{\rm div}\bar{E}_t^\lambda,\lambda\nabla{\rm div}\bar{E}^\lambda,\lambda\nabla{\rm div}\bar{E}_t^\lambda) \parallel^2(t=0).\label{eq:txg4-1} \end{matrix}$
不等式(4.6)是一个$\lambda$ -加权的Growall型熵积分不等式, 因此有下面的结论.
(4.8) $\begin{equation} \tilde{\Gamma^\lambda}(t=0)\leq M \lambda, \label{eq:init-s} \end{equation}$
其中$M$ 是独立于$\lambda$ 的正常数. 那么对于任意的$T\in (0,T_{\max})$ , $T_{\max}\leq\infty$ , 存在正常数$\lambda_0\ll1$ , 使得对于任意的$\lambda\leq \lambda_0,\delta\in (0,1), 0\leq t\leq T$ 不等式
(4.9) $\begin{equation} \Gamma^\lambda(t)\leq M \lambda^{1-\delta}\label{eq:con-res} \end{equation}$
注4.1 此定理的证明类似于文献[引理10], 此处省略其证明.
按照初值的假设, 易知(4.8)式成立, 应用定理3.5, 知不等式(4.9)成立. 由(4.9)式, 可得不等式(2.24). 定理得证.
参考文献
View Option
[2]
Rubinstein I . Electro-Diffusion of Ions
Philadelphia: Siam , 1990
[本文引用: 1]
[4]
Wang S , Jiang L M , Liu C D . Quasi-neutral limit and the boundary layer problem of Planck Nernst Poisson Navier Stokes equations for electro hydrodynamics
J Diff Equations , 2009 , 267 : 3475 -3523
DOI:10.1016/j.jde.2019.04.011
URL
[本文引用: 3]
[5]
Wang S , Jiang L M . Quasi-neutral limit and the boundary layer problem of the electro diffusion model arising in electro hydrodynamics
Nonlinear Anal: RWA , 2021 , 59 : 103266
DOI:10.1016/j.nonrwa.2020.103266
URL
[本文引用: 1]
[6]
Yang J W , Ju Q C . Convergence of the quantum Navier-Stokes-Poisson equations to the incompressible Euler equations for general initial data
Nonlinear Anal: RWA , 2015 , 23 : 148 -159
DOI:10.1016/j.nonrwa.2014.12.003
URL
[本文引用: 1]
[9]
Alì D , Bini D , Rionero S . Existence and relaxation limit for smooth solution to the Euler-Poisson model for semiconductors
Siam J Math Anal , 2000 , 32 : 572 -587
DOI:10.1137/S0036141099355174
URL
[本文引用: 1]
[11]
Gasser I , Levermore C D , Markowich P , Shmeiser C . The initial time layer problem and the quasi-neutral limit in the semiconductor drift-diffusion model
European J Appl Math , 2001 , 12 : 497 -512
DOI:10.1017/S0956792501004533
URL
[本文引用: 1]
[12]
Guo Y , Strauss W . Stability of semiconductor states with insulating and contact boundary conditions
Arch Rat Mech and Anal , 2006 , 179 : 1 -30
DOI:10.1007/s00205-005-0369-2
URL
[本文引用: 1]
[13]
Hsiao L , Li F C , Wang S . Coupled quasi-neutral and inviscid limit of the Vlasov-Poisson-Fokker-Planck system
Commun Pure Appl Anal , 2008 , 7 : 579 -589
[本文引用: 1]
[14]
Jüngel A . Peng Y J . Ahierarchy of hydrodynamic models for plasmas: Quasi-neutral limits in the drift-diffusion equations
Asymptot Anal , 2001 , 28 : 49 -73
[本文引用: 1]
[15]
Roubicek T . Nonlinear Partial Differential Equations with Applications . Basel : Birkhauser Verlag , 2005
[本文引用: 1]
[16]
Suzuki M . Asymptotic stability of stationary solutions to the Euler-Poisson equations arising in plasma physics
Kinet Relat Models , 2011 , 4 : 569 -588
DOI:10.3934/krm.2011.4.569
URL
[本文引用: 1]
[17]
Temam R . Navier-Stokes Equations Theory and Numerical Analysis . New York : North-Holland , 1977
[本文引用: 1]
[18]
Wang S . Quasi-neutral limit of Euler-Poisson system with and without viscosity
Comm Part Diff Equations , 2004 , 29 : 419 -456
DOI:10.1081/PDE-120030403
URL
[本文引用: 1]
[19]
Wang S . Quasi-neutral limit of the multi-dimensional drift-diffusion-Poisson model for semiconductor with pn-junctions
Math Models Methods Appl Sci , 2006 , 16 : 737 -757
[本文引用: 1]
[20]
Hsiao L , Wang S . Quasi-neureal limit of a time dependent drift-diffusion-Poisson model for p-n junction semiconductor devices
J Diff Eqns , 2006 , 225 : 411 -439
DOI:10.1016/j.jde.2006.01.022
URL
Invariant regions for the Nernst-Planck equations
1
1998
... 该文研究了迁移率互异的可压电扩散模型PNPNS(Planck Nernest Poisson Navier Stokes) 的拟中性极限[1 -2 ] . 该系统为 ...
Electro-Diffusion of Ions
1
1990
... 该文研究了迁移率互异的可压电扩散模型PNPNS(Planck Nernest Poisson Navier Stokes) 的拟中性极限[1 -2 ] . 该系统为 ...
Quasi-neutral limit of the electro-diffusion model arising in electrohydrodynamics
1
2009
... 据作者所知, 关于此系统的一些结论. 在掺杂函数光滑的假设下, Li[3 ] 证明了不可压电解液中电扩散方程的拟中性极限. Wang等[4 ] 研究了三维空间下带有不同迁移率电解液中不可压电扩散模型的拟中性极限和边界层问题. Wang等[5 ] 研究了电解液中不可压电扩散模型的初始层问题. Yang等[6 ] 研究了一般初值下量子 Navier-Stokes-Poisson 方程到不可压 Euler方程的收敛性问题. Liu等[7 ] 研究了三维空间中 Prandtl 非正定性. 拟中性问题被国内外专家学者广泛关注,并且有很多有意义的结论, 比如文献[8 ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ -19 ] 以及其参考文献. ...
Quasi-neutral limit and the boundary layer problem of Planck Nernst Poisson Navier Stokes equations for electro hydrodynamics
3
2009
... 据作者所知, 关于此系统的一些结论. 在掺杂函数光滑的假设下, Li[3 ] 证明了不可压电解液中电扩散方程的拟中性极限. Wang等[4 ] 研究了三维空间下带有不同迁移率电解液中不可压电扩散模型的拟中性极限和边界层问题. Wang等[5 ] 研究了电解液中不可压电扩散模型的初始层问题. Yang等[6 ] 研究了一般初值下量子 Navier-Stokes-Poisson 方程到不可压 Euler方程的收敛性问题. Liu等[7 ] 研究了三维空间中 Prandtl 非正定性. 拟中性问题被国内外专家学者广泛关注,并且有很多有意义的结论, 比如文献[8 ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ -19 ] 以及其参考文献. ...
... 对于非线性项, 应用Cauchy-Schwarz不等式, Sobolev嵌入不等式, 见文献[4 ], 有 ...
... 估计方程(3.29)右端各项. 对于线性项, 应用Cauchy-Schwarz不等式, Sobolev嵌入定理, 见文献[4 ], 线性项可由(3.30)式控制 ...
Quasi-neutral limit and the boundary layer problem of the electro diffusion model arising in electro hydrodynamics
1
2021
... 据作者所知, 关于此系统的一些结论. 在掺杂函数光滑的假设下, Li[3 ] 证明了不可压电解液中电扩散方程的拟中性极限. Wang等[4 ] 研究了三维空间下带有不同迁移率电解液中不可压电扩散模型的拟中性极限和边界层问题. Wang等[5 ] 研究了电解液中不可压电扩散模型的初始层问题. Yang等[6 ] 研究了一般初值下量子 Navier-Stokes-Poisson 方程到不可压 Euler方程的收敛性问题. Liu等[7 ] 研究了三维空间中 Prandtl 非正定性. 拟中性问题被国内外专家学者广泛关注,并且有很多有意义的结论, 比如文献[8 ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ -19 ] 以及其参考文献. ...
Convergence of the quantum Navier-Stokes-Poisson equations to the incompressible Euler equations for general initial data
1
2015
... 据作者所知, 关于此系统的一些结论. 在掺杂函数光滑的假设下, Li[3 ] 证明了不可压电解液中电扩散方程的拟中性极限. Wang等[4 ] 研究了三维空间下带有不同迁移率电解液中不可压电扩散模型的拟中性极限和边界层问题. Wang等[5 ] 研究了电解液中不可压电扩散模型的初始层问题. Yang等[6 ] 研究了一般初值下量子 Navier-Stokes-Poisson 方程到不可压 Euler方程的收敛性问题. Liu等[7 ] 研究了三维空间中 Prandtl 非正定性. 拟中性问题被国内外专家学者广泛关注,并且有很多有意义的结论, 比如文献[8 ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ -19 ] 以及其参考文献. ...
On the ill-posedness of the prandtl equations in three-dimensional space
1
2016
... 据作者所知, 关于此系统的一些结论. 在掺杂函数光滑的假设下, Li[3 ] 证明了不可压电解液中电扩散方程的拟中性极限. Wang等[4 ] 研究了三维空间下带有不同迁移率电解液中不可压电扩散模型的拟中性极限和边界层问题. Wang等[5 ] 研究了电解液中不可压电扩散模型的初始层问题. Yang等[6 ] 研究了一般初值下量子 Navier-Stokes-Poisson 方程到不可压 Euler方程的收敛性问题. Liu等[7 ] 研究了三维空间中 Prandtl 非正定性. 拟中性问题被国内外专家学者广泛关注,并且有很多有意义的结论, 比如文献[8 ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ -19 ] 以及其参考文献. ...
Convergence of the Vlasov-Poisson system to the incompressible Euler equations
1
2000
... 据作者所知, 关于此系统的一些结论. 在掺杂函数光滑的假设下, Li[3 ] 证明了不可压电解液中电扩散方程的拟中性极限. Wang等[4 ] 研究了三维空间下带有不同迁移率电解液中不可压电扩散模型的拟中性极限和边界层问题. Wang等[5 ] 研究了电解液中不可压电扩散模型的初始层问题. Yang等[6 ] 研究了一般初值下量子 Navier-Stokes-Poisson 方程到不可压 Euler方程的收敛性问题. Liu等[7 ] 研究了三维空间中 Prandtl 非正定性. 拟中性问题被国内外专家学者广泛关注,并且有很多有意义的结论, 比如文献[8 ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ -19 ] 以及其参考文献. ...
Existence and relaxation limit for smooth solution to the Euler-Poisson model for semiconductors
1
2000
... 据作者所知, 关于此系统的一些结论. 在掺杂函数光滑的假设下, Li[3 ] 证明了不可压电解液中电扩散方程的拟中性极限. Wang等[4 ] 研究了三维空间下带有不同迁移率电解液中不可压电扩散模型的拟中性极限和边界层问题. Wang等[5 ] 研究了电解液中不可压电扩散模型的初始层问题. Yang等[6 ] 研究了一般初值下量子 Navier-Stokes-Poisson 方程到不可压 Euler方程的收敛性问题. Liu等[7 ] 研究了三维空间中 Prandtl 非正定性. 拟中性问题被国内外专家学者广泛关注,并且有很多有意义的结论, 比如文献[8 ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ -19 ] 以及其参考文献. ...
Global smooth solutions to the multi-dimensional hydrodynamic model for two-carrier plasmas
1
2003
... 据作者所知, 关于此系统的一些结论. 在掺杂函数光滑的假设下, Li[3 ] 证明了不可压电解液中电扩散方程的拟中性极限. Wang等[4 ] 研究了三维空间下带有不同迁移率电解液中不可压电扩散模型的拟中性极限和边界层问题. Wang等[5 ] 研究了电解液中不可压电扩散模型的初始层问题. Yang等[6 ] 研究了一般初值下量子 Navier-Stokes-Poisson 方程到不可压 Euler方程的收敛性问题. Liu等[7 ] 研究了三维空间中 Prandtl 非正定性. 拟中性问题被国内外专家学者广泛关注,并且有很多有意义的结论, 比如文献[8 ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ -19 ] 以及其参考文献. ...
The initial time layer problem and the quasi-neutral limit in the semiconductor drift-diffusion model
1
2001
... 据作者所知, 关于此系统的一些结论. 在掺杂函数光滑的假设下, Li[3 ] 证明了不可压电解液中电扩散方程的拟中性极限. Wang等[4 ] 研究了三维空间下带有不同迁移率电解液中不可压电扩散模型的拟中性极限和边界层问题. Wang等[5 ] 研究了电解液中不可压电扩散模型的初始层问题. Yang等[6 ] 研究了一般初值下量子 Navier-Stokes-Poisson 方程到不可压 Euler方程的收敛性问题. Liu等[7 ] 研究了三维空间中 Prandtl 非正定性. 拟中性问题被国内外专家学者广泛关注,并且有很多有意义的结论, 比如文献[8 ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ -19 ] 以及其参考文献. ...
Stability of semiconductor states with insulating and contact boundary conditions
1
2006
... 据作者所知, 关于此系统的一些结论. 在掺杂函数光滑的假设下, Li[3 ] 证明了不可压电解液中电扩散方程的拟中性极限. Wang等[4 ] 研究了三维空间下带有不同迁移率电解液中不可压电扩散模型的拟中性极限和边界层问题. Wang等[5 ] 研究了电解液中不可压电扩散模型的初始层问题. Yang等[6 ] 研究了一般初值下量子 Navier-Stokes-Poisson 方程到不可压 Euler方程的收敛性问题. Liu等[7 ] 研究了三维空间中 Prandtl 非正定性. 拟中性问题被国内外专家学者广泛关注,并且有很多有意义的结论, 比如文献[8 ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ -19 ] 以及其参考文献. ...
Coupled quasi-neutral and inviscid limit of the Vlasov-Poisson-Fokker-Planck system
1
2008
... 据作者所知, 关于此系统的一些结论. 在掺杂函数光滑的假设下, Li[3 ] 证明了不可压电解液中电扩散方程的拟中性极限. Wang等[4 ] 研究了三维空间下带有不同迁移率电解液中不可压电扩散模型的拟中性极限和边界层问题. Wang等[5 ] 研究了电解液中不可压电扩散模型的初始层问题. Yang等[6 ] 研究了一般初值下量子 Navier-Stokes-Poisson 方程到不可压 Euler方程的收敛性问题. Liu等[7 ] 研究了三维空间中 Prandtl 非正定性. 拟中性问题被国内外专家学者广泛关注,并且有很多有意义的结论, 比如文献[8 ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ -19 ] 以及其参考文献. ...
Ahierarchy of hydrodynamic models for plasmas: Quasi-neutral limits in the drift-diffusion equations
1
2001
... 据作者所知, 关于此系统的一些结论. 在掺杂函数光滑的假设下, Li[3 ] 证明了不可压电解液中电扩散方程的拟中性极限. Wang等[4 ] 研究了三维空间下带有不同迁移率电解液中不可压电扩散模型的拟中性极限和边界层问题. Wang等[5 ] 研究了电解液中不可压电扩散模型的初始层问题. Yang等[6 ] 研究了一般初值下量子 Navier-Stokes-Poisson 方程到不可压 Euler方程的收敛性问题. Liu等[7 ] 研究了三维空间中 Prandtl 非正定性. 拟中性问题被国内外专家学者广泛关注,并且有很多有意义的结论, 比如文献[8 ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ -19 ] 以及其参考文献. ...
1
2005
... 据作者所知, 关于此系统的一些结论. 在掺杂函数光滑的假设下, Li[3 ] 证明了不可压电解液中电扩散方程的拟中性极限. Wang等[4 ] 研究了三维空间下带有不同迁移率电解液中不可压电扩散模型的拟中性极限和边界层问题. Wang等[5 ] 研究了电解液中不可压电扩散模型的初始层问题. Yang等[6 ] 研究了一般初值下量子 Navier-Stokes-Poisson 方程到不可压 Euler方程的收敛性问题. Liu等[7 ] 研究了三维空间中 Prandtl 非正定性. 拟中性问题被国内外专家学者广泛关注,并且有很多有意义的结论, 比如文献[8 ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ -19 ] 以及其参考文献. ...
Asymptotic stability of stationary solutions to the Euler-Poisson equations arising in plasma physics
1
2011
... 据作者所知, 关于此系统的一些结论. 在掺杂函数光滑的假设下, Li[3 ] 证明了不可压电解液中电扩散方程的拟中性极限. Wang等[4 ] 研究了三维空间下带有不同迁移率电解液中不可压电扩散模型的拟中性极限和边界层问题. Wang等[5 ] 研究了电解液中不可压电扩散模型的初始层问题. Yang等[6 ] 研究了一般初值下量子 Navier-Stokes-Poisson 方程到不可压 Euler方程的收敛性问题. Liu等[7 ] 研究了三维空间中 Prandtl 非正定性. 拟中性问题被国内外专家学者广泛关注,并且有很多有意义的结论, 比如文献[8 ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ -19 ] 以及其参考文献. ...
1
1977
... 据作者所知, 关于此系统的一些结论. 在掺杂函数光滑的假设下, Li[3 ] 证明了不可压电解液中电扩散方程的拟中性极限. Wang等[4 ] 研究了三维空间下带有不同迁移率电解液中不可压电扩散模型的拟中性极限和边界层问题. Wang等[5 ] 研究了电解液中不可压电扩散模型的初始层问题. Yang等[6 ] 研究了一般初值下量子 Navier-Stokes-Poisson 方程到不可压 Euler方程的收敛性问题. Liu等[7 ] 研究了三维空间中 Prandtl 非正定性. 拟中性问题被国内外专家学者广泛关注,并且有很多有意义的结论, 比如文献[8 ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ -19 ] 以及其参考文献. ...
Quasi-neutral limit of Euler-Poisson system with and without viscosity
1
2004
... 据作者所知, 关于此系统的一些结论. 在掺杂函数光滑的假设下, Li[3 ] 证明了不可压电解液中电扩散方程的拟中性极限. Wang等[4 ] 研究了三维空间下带有不同迁移率电解液中不可压电扩散模型的拟中性极限和边界层问题. Wang等[5 ] 研究了电解液中不可压电扩散模型的初始层问题. Yang等[6 ] 研究了一般初值下量子 Navier-Stokes-Poisson 方程到不可压 Euler方程的收敛性问题. Liu等[7 ] 研究了三维空间中 Prandtl 非正定性. 拟中性问题被国内外专家学者广泛关注,并且有很多有意义的结论, 比如文献[8 ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ -19 ] 以及其参考文献. ...
Quasi-neutral limit of the multi-dimensional drift-diffusion-Poisson model for semiconductor with pn-junctions
1
2006
... 据作者所知, 关于此系统的一些结论. 在掺杂函数光滑的假设下, Li[3 ] 证明了不可压电解液中电扩散方程的拟中性极限. Wang等[4 ] 研究了三维空间下带有不同迁移率电解液中不可压电扩散模型的拟中性极限和边界层问题. Wang等[5 ] 研究了电解液中不可压电扩散模型的初始层问题. Yang等[6 ] 研究了一般初值下量子 Navier-Stokes-Poisson 方程到不可压 Euler方程的收敛性问题. Liu等[7 ] 研究了三维空间中 Prandtl 非正定性. 拟中性问题被国内外专家学者广泛关注,并且有很多有意义的结论, 比如文献[8 ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ ⇓ -19 ] 以及其参考文献. ...
Quasi-neureal limit of a time dependent drift-diffusion-Poisson model for p-n junction semiconductor devices
2006