删除或更新信息,请邮件至freekaoyan#163.com(#换成@)

一个联系特殊函数的多参数Hilbert型积分不等式

本站小编 Free考研考试/2021-12-25

黄琳1, 刘琼2
1. 长沙师范学院师范预科部, 长沙 410100;
2. 邵阳学院理学与信息科学系, 湖南 邵阳 422000
2016年01月08日 收稿; 2016年03月26日 收修改稿
基金项目: 国家自然科学基金(11171280)和湖南省教育厅科学研究项目(10C1186)资助
通信作者: 黄琳?E-mail:13787317290@163.com

摘要: 利用权函数方法和实分析及泛函技巧,引入一些特殊函数联合刻划常数因子,建立一个多参数Hilbert型积分不等式,考虑它的等价式,证明它们的常数因子是最佳的.作为应用,通过选取特殊的参数值,得到一些有意义的结果.
关键词: Hilbert型积分不等式权函数最佳常数因子特殊函数
A multi-parameter Hilbert-type integral inequality related to special functions
HUANG Lin1, LIU Qiong2
1. Preparatory Department of Junior Education, Changsha Normal University, Changsha 410100, China;
2. Department of Science and Information, Shaoyang University, Shaoyang 422000, Hunan, China


Abstract: By using the method of weight function and the techniques of real analysis and functional analysis and by introducing some special functions to jointly score the constant factor, a Hilbert-type integral inequality with multi-parameters is given. Its equivalent form is considered, and their constant factor is proved to be the best. Some meaningful results are obtained by choosing the special parameter values.
Key words: Hilbert-type integral inequalityweight functionthe best constant factorspecial function
为后面的叙述方便, 设θ(x)(>0)为可测函数, ρ≥1, 定义如下函数空间:
$\begin{array}{l}{L^\rho }\left( {0, \infty } \right):\; = \left\{ {{{\left\| h \right\|}_\rho }:\; = } \right.\\\;\;\;\;\left. {{{\left\{ {\int_0^\infty {{{\left| {h\left( x \right)} \right|}^\rho }{\rm{d}}x} } \right\}}^{\frac{1}{\rho }}} < \infty } \right\}, \end{array}$

$\begin{array}{l}L_\theta ^\rho \left( {0, \infty } \right):\; = \left\{ {{{\left\| h \right\|}_{\rho, \theta }}:\; = } \right.\\\;\;\;\left. {{{\left\{ {\int_0^\infty {\theta \left( x \right){{\left| {h\left( x \right)} \right|}^\rho }{\rm{d}}x} } \right\}}^{\frac{1}{\rho }}} < \infty } \right\}.\end{array}$
$f, g \ge 0, f, g, \in {L^2}\left( {0, \infty } \right), {\left\| f \right\|_2}, {\left\| g \right\|_2} > 0$, 则有下面的Hilbert积分不等式[1]
$\int_0^\infty {\int_0^\infty {\frac{{f\left( x \right)g\left( y \right)}}{{x + y}}{\rm{d}}x{\rm{d}}y < {\rm{\pi }}{{\left\| f \right\|}_2}{{\left\| g \right\|}_2}, } } $ (1)
这里的常数因子π是最佳值.在与式(1)相同的条件下, 还有下面基本Hilbert型积分不等式[2]:
$\int_0^\infty {\int_0^\infty {\frac{{\left| {\ln \frac{x}{y}} \right|f\left( x \right)g\left( y \right)}}{{x + y}}{\rm{d}}x{\rm{d}}y < {c_0}{{\left\| f \right\|}_2}{{\left\| g \right\|}_2}, } } $ (2)
这里的常数因子${c_0}\left( { = \sum\limits_{k = 1}^\infty {\frac{{8{{\left( {-1} \right)}^{k-1}}}}{{{{\left( {2k-1} \right)}^2}}} = 7.327\;{7^ + }} } \right)$是最佳值.近年来, 人们在Hilbert型积分不等式研究中的主要成就:一方面是将以前的基本核进行组合, 得到一些混合核的积分不等式, 同时进行参量化研究, 综合、推广和改进已有结果[3-6].另一方面, 构造一些新的积分核, 发现新的Hilbert型积分不等式[7-10].这些所获得的不等式在分析学和偏微分方程理论等领域有重要应用.本文引入Γ-函数、推广的ζ-函数等刻划常数因子, 利用权函数方法和实分析的技巧, 建立一个联系特殊函数的多参数Hilbert型积分不等式, 给出它的等价式, 证明了它们的常数因子是最佳的, 并通过选取特殊参数值, 得到一些有意义的结果.
1 有关引理本文将用到以下特殊函数[11]:
Γ-函数:
$\Gamma \left( z \right) = \int_0^\infty {{{\rm{e}}^{-u}}{u^{z-1}}{\rm{d}}u, \left( {z > 0} \right)}, $ (3)
黎曼ζ-函数:
$\zeta \left( s \right) = \sum\limits_{k = 1}^\infty {\frac{1}{{{k^s}}}\left( {{\mathop{\rm Re}\nolimits} \left( s \right) > 1} \right)}, $ (4)
推广的ζ-函数:
$\zeta \left( {s, a} \right) = \sum\limits_{k = 0}^\infty {\frac{1}{{{{\left( {k + a} \right)}^s}}}}, $ (5)
这里Re(s)>1, a不等于零和负整数.显然, ζ(s, 1)=ζ(s).
引理1.1 ??设Re(s)>1, $\frac{a}{2}$$\frac{{a + 1}}{2}$均不等于零和负整数, 则有求和公式
$\sum\limits_{k = 0}^\infty {\frac{{{{\left( {- 1} \right)}^k}}}{{{{\left( {k + a} \right)}^s}}}} = \frac{1}{{{2^s}}}\left[{\zeta \left( {s, \frac{a}{2}} \right)-\zeta \left( {s, \frac{{a + 1}}{2}} \right)} \right].$ (6)
证明
$\begin{array}{l}\sum\limits_{k = 0}^\infty {\frac{{{{\left( {- 1} \right)}^k}}}{{{{\left( {k + a} \right)}^s}}} = \sum\limits_{k = 0}^\infty {\frac{1}{{{{\left( {2k + a} \right)}^s}}}- \sum\limits_{k = 0}^\infty {\frac{1}{{{{\left( {2k + 1 + a} \right)}^s}}}} } } \\ = \frac{1}{{{2^s}}}\left[{\sum\limits_{k = 0}^\infty {\frac{1}{{{{\left( {k + \frac{a}{2}} \right)}^s}}}}-\sum\limits_{k = 0}^\infty {\frac{1}{{{{\left( {k + \frac{{a + 1}}{2}} \right)}^s}}}} } \right]\\ = \frac{1}{{{2^s}}}\left[{\zeta \left( {s, \frac{a}{2}} \right)-\zeta \left( {s, \frac{{a + 1}}{2}} \right)} \right].\end{array}$
引理1.2 ??设$p > 1, \frac{1}{p} + \frac{1}{q} = 1, \alpha > 0, \frac{{\beta + 1}}{4}$$\frac{{\beta + 3}}{4}$均不为零和负整数, 定义如下权函数:
$\begin{array}{l}\omega \left( {\alpha, \beta, x} \right) = \int_0^\infty {\frac{{{{\left| {\ln \frac{x}{y}} \right|}^\alpha }{{\left( {\min \left\{ {x, y} \right\}} \right)}^\beta }}}{{x + y}}} \\\frac{{{y^{-\frac{{\beta + 1}}{2}}}}}{{{x^{-\frac{{p\left( {\beta + 1} \right)}}{{2q}}}}}}{\rm{d}}y{\rm{, }}x \in \left( {0, + \infty } \right), \\\omega \left( {\alpha, \beta, y} \right) = \int_0^\infty {\frac{{{{\left| {\ln \frac{x}{y}} \right|}^\alpha }{{\left( {\min \left\{ {x, y} \right\}} \right)}^\beta }}}{{x + y}}} \\\frac{{{y^{-\frac{{\beta + 1}}{2}}}}}{{{x^{ - \frac{{q\left( {\beta + 1} \right)}}{{2p}}}}}}{\rm{d}}x, y \in \left( {0, + \infty } \right), \;则\end{array}$
$\begin{array}{l}\omega \left( {\alpha, \beta, x} \right) = C\left( {\alpha, \beta } \right){x^{\frac{{p\left( {\beta + 1} \right)}}{2}-1}}, \\\omega \left( {\alpha, \beta, y} \right) = C\left( {\alpha, \beta } \right){y^{\frac{{q\left( {\beta + 1} \right)}}{2}-1}}, \end{array}$
其中
$\begin{array}{l}C\left( {\alpha, \beta } \right) = \frac{1}{{{2^\alpha }}}\left[{\zeta \left( {\alpha + 1, \frac{{\beta + 1}}{4}} \right)-} \right.\\\;\;\;\;\;\;\left. {\zeta \left( {\alpha + 1, \frac{{\beta + 3}}{4}} \right)} \right]\Gamma \left( {\alpha + 1} \right).\end{array}$ (7)
证明$\frac{y}{x} = u$, 由Fubini定理12和引理1.1有
$\begin{array}{l}\omega \left( {\alpha, \beta, x} \right) = \int_0^\infty {\frac{{{{\left| {\ln \frac{x}{y}} \right|}^\alpha }{{\left( {\min \left\{ {x, y} \right\}} \right)}^\beta }}}{{x + y}}} \frac{{{y^{- \frac{{\beta + 1}}{2}}}}}{{{x^{- \frac{{p\left( {\beta + 1} \right)}}{{2q}}}}}}{\rm{d}}y\\ = {x^{\frac{{p\left( {\beta + 1} \right)}}{2}- 1}}\int_0^\infty {\frac{{{{\left| {\ln \;u} \right|}^\alpha }{{\left( {\min \left\{ {1, u} \right\}} \right)}^\beta }{u^{ - \frac{{\beta + 1}}{2}}}}}{{1 + u}}{\rm{d}}u} \\ = 2{x^{\frac{{p\left( {\beta + 1} \right)}}{2} - 1}}\int_0^1 {\frac{{{{\left| {\ln u} \right|}^\alpha }{u^{\frac{{\beta - 1}}{2}}}}}{{1 + u}}{\rm{d}}u} \\ = 2{x^{\frac{{p\left( {\beta + 1} \right)}}{2} - 1}}\int_0^\infty {\frac{{{{\rm{e}}^{ - \left( {\frac{{\beta + 1}}{2}} \right)t}}{t^\alpha }}}{{1 + {{\rm{e}}^{ - t}}}}{\rm{d}}t} \\ = 2{x^{\frac{{p\left( {\beta + 1} \right)}}{2} - 1}}\sum\limits_{k = 0}^\infty {{{\left( { - 1} \right)}^k}\int_0^\infty {{{\rm{e}}^{ - \left( {k + \frac{{\beta + 1}}{2}} \right)t}}} {t^\alpha }{\rm{d}}t} \\ = 2{x^{\frac{{p\left( {\beta + 1} \right)}}{2} - 1}}\sum\limits_{k = 0}^\infty {\frac{{{{\left( { - 1} \right)}^k}}}{{{{\left( {k + \frac{{\beta + 1}}{2}} \right)}^{\alpha + 1}}}}\int_0^\infty {{{\rm{e}}^{ - t}}{t^\alpha }{\rm{d}}t} } \\ = \frac{1}{{{2^\alpha }}}\left[{\zeta \left( {\alpha + 1, \frac{{\beta + 1}}{4}} \right)-\zeta \left( {\alpha + 1, \frac{{\beta + 3}}{4}} \right)} \right] \times \\\Gamma \left( {\alpha + 1} \right){x^{\frac{{p\left( {\beta + 1} \right)}}{2} -1}} = C\left( {\alpha, \beta } \right){x^{\frac{{p\left( {\beta + 1} \right)}}{2} -1}}.\end{array}$
同理可证$\omega \left( {\alpha, \beta, y} \right) = C\left( {\alpha, \beta } \right){y^{\frac{{q\left( {\beta + 1} \right)}}{2}-1}}$.
引理1.3 ??设$p > 1, \frac{1}{p} + \frac{1}{q} = 1, \alpha > 0, \beta >- 1, \varepsilon + \sqrt[3]{\varepsilon } < \frac{{q\left( {\beta + 1} \right)}}{2}$, 且0 < ε可以充分地小, 定义如下函数:
$\begin{array}{l}\tilde f\left( x \right) = \left\{ \begin{array}{l}0, \;\;\;\;\;\;\;\;\;\;\;\;\;\;\;x \in \left( {0, 1} \right)\\{x^{\frac{{- \frac{{p\left( {\beta + 1} \right)}}{2}- \varepsilon }}{p}}}, \;\;\;\;x \in \left[{1, \infty } \right)\end{array} \right., \\\tilde g\left( y \right) = \left\{ \begin{array}{l}0, \;\;\;\;\;\;\;\;\;\;\;\;\;\;\;y \in \left( {0, 1} \right)\\{y^{\frac{{- \frac{{q\left( {\beta + 1} \right)}}{2}- \varepsilon }}{q}}}, \;\;\;\;y \in \left[{1, \infty } \right)\end{array} \right., \end{array}$
则有
$\begin{array}{l}\tilde J\varepsilon = {\left[{\int_0^\infty {{x^{\frac{{p\left( {\beta-1} \right)}}{2}-1}}} {{\tilde f}^p}\left( x \right){\rm{d}}x} \right]^{\frac{1}{p}}} \times \\\;\;\;\;\;\;\;\;{\left[{\int_0^\infty {{y^{\frac{{q\left( {\beta + 1} \right)}}{2}-1}}{{\tilde g}^q}\left( y \right){\rm{d}}y} } \right]^{\frac{1}{q}}}\varepsilon = 1, \end{array}$ (8)
$\begin{array}{l}\tilde I\varepsilon = \varepsilon \int_0^\infty {\int_0^\infty {\frac{{{{\left| {\ln \frac{x}{y}} \right|}^\alpha }{{\left( {\min \left\{ {x, y} \right\}} \right)}^\beta }\tilde f\left( x \right)\tilde g\left( y \right)}}{{x + y}}} } {\rm{d}}x{\rm{d}}y\\\;\;\;\;\;\;\;\; > C\left( {\alpha, \beta } \right)\left( {1-o\left( 1 \right)} \right)\left( {\varepsilon \to {0^ + }} \right).\end{array}$ (9)
证明 ??容易得到
$\begin{array}{l}\tilde J\varepsilon = {\left[{\int_0^\infty {{x^{\frac{{p\left( {\beta-1} \right)}}{2}-1}}} {{\tilde f}^p}\left( x \right){\rm{d}}x} \right]^{\frac{1}{p}}} \times \\\;\;\;\;\;\;\;\;{\left[{\int_0^\infty {{y^{\frac{{q\left( {\beta + 1} \right)}}{2}-1}}{{\tilde g}^q}\left( y \right){\rm{d}}y} } \right]^{\frac{1}{q}}}\varepsilon \\\;\;\;\;\;\; = {\left[{\int_1^\infty {{x^{-\left( {1 + \varepsilon } \right)}}{\rm{d}}x} } \right]^{\frac{1}{p}}}{\left[{\int_1^\infty {{y^{-\left( {1 + \varepsilon } \right)}}{\rm{d}}y} } \right]^{\frac{1}{q}}}\varepsilon \\\;\;\;\;\;\; = 1.\end{array}$
因为$F\left( t \right) = \frac{{{t^{\frac{{\beta + 1}}{2}- \frac{{\varepsilon + \sqrt[3]{\varepsilon }}}{q}}}{{\left| {\ln t} \right|}^\alpha }}}{{1 + t}}$在(0, 1]内连续, 且用洛比达法则得$\mathop {\lim }\limits_{t \to {0^ + }} F\left( t \right) = \mathop {\lim }\limits_{t \to {0^ + }} \frac{{{t^{\frac{{\beta + 1}}{2}- \frac{{\varepsilon + \sqrt[3]{\varepsilon }}}{q}}}{{\left| {\ln \;t} \right|}^\alpha }}}{{1 + t}} = 0$, 故存在M>0, 使F(t)≤M, 则有
$\begin{array}{l}\tilde I\varepsilon = \varepsilon \int_0^\infty {\int_0^\infty {\frac{{{{\left| {\ln \frac{x}{y}} \right|}^\alpha }{{\left( {\min \;\left\{ {x, y} \right\}} \right)}^\beta }\tilde f\left( x \right)\tilde g\left( y \right)}}{{x + y}}} {\rm{d}}x{\rm{d}}y} \\ = \varepsilon \int_1^\infty {{x^{\frac{{- \frac{{p\left( {\beta + 1} \right)}}{2}- \varepsilon }}{p}}}{\rm{d}}x\left[{\int_1^\infty {\frac{{{{\left| {\ln \frac{x}{y}} \right|}^\alpha }{{\left( {\min \left\{ {x, y} \right\}} \right)}^\beta }}}{{x + y}} \times } } \right.} \\\;\;\;\;\;\;\;\left. {{y^{\frac{{-\frac{{q\left( {\beta + 1} \right)}}{2}-\varepsilon }}{q}}}{\rm{d}}y} \right]\\\;\; = \varepsilon \int_1^\infty {{x^{ - 1 - \varepsilon }}{\rm{d}}x\left[{\int_0^1 {\frac{{{{\left| {\ln \;t} \right|}^\alpha }{t^{\frac{{\beta-1}}{2}-\frac{\varepsilon }{q}}}}}{{1 + t}}} } \right.} {\rm{d}}t + \\\;\;\;\left. {\int_0^1 {\frac{{{{\left| {\ln \;t} \right|}^\alpha }{t^{\frac{{\beta-1}}{2} + \frac{\varepsilon }{q}}}}}{{1 + t}}{\rm{d}}t - \int_0^{{x^{ - 1}}} {\frac{{{{\left| {\ln \;t} \right|}^\alpha }{t^{\frac{{\beta - 1}}{2} - \frac{\varepsilon }{q}}}}}{{1 + t}}{\rm{d}}t} } } \right]\\\;\;=\int_{0}^{\infty }{\frac{\left( {{\text{e}}^{-\ \frac{u\left( \beta +1 \right)}{2}+\frac{\varepsilon }{q}}}\text{+}{{\text{e}}^{-\ \frac{u\left( \beta +1 \right)}{2}-\frac{\varepsilon }{q}}} \right){{u}^{\alpha }}}{1+{{\text{e}}^{-u}}}}\text{d}u-\\\;\;\varepsilon \int_1^\infty {{x^{ - 1 - \varepsilon }}{\rm{d}}x\int_0^{{x^{ - 1}}} {\frac{{{{\left| {\ln \;t} \right|}^\alpha }{t^{\frac{{\beta - 1}}{2} - \frac{\varepsilon }{q}}}}}{{1 + t}}{\rm{d}}t} } \\\;\; = \sum\limits_{k = 0}^\infty {\frac{{{{\left( { - 1} \right)}^k}}}{{{{\left( {k + \frac{{\beta + 1}}{2} + \frac{\varepsilon }{q}} \right)}^{\alpha + 1}}}}} \int_0^\infty {{{\rm{e}}^{ - u}}{u^\alpha }{\rm{d}}u + } \\\;\;\sum\limits_{k = 0}^\infty {\frac{{{{\left( { - 1} \right)}^k}}}{{{{\left( {k + \frac{{\beta + 1}}{2} - \frac{\varepsilon }{q}} \right)}^{\alpha + 1}}}}\int_0^\infty {{{\rm{e}}^{ - u}}{u^\alpha }{\rm{d}}u - } } \\\;\;\;\;\varepsilon \int_1^\infty {{x^{ - 1 - \varepsilon }}{\rm{d}}x\int_0^{{x^{ - 1}}} {\frac{{{{\left| {\ln \;t} \right|}^\alpha }{t^{\frac{{\beta - 1}}{2} - \frac{\varepsilon }{q}}}}}{{1 + t}}{\rm{d}}t} } \\ > C\left( {\alpha, \beta } \right) + {o_1}\left( 1 \right) - \\\;\;\;M\varepsilon \int_1^\infty {{x^{ - 1}}{\rm{d}}x\int_0^{{x^{ - 1}}} {{t^{ - 1 + \frac{{\sqrt[3]{\varepsilon }}}{q}}}{\rm{d}}t} } \\ = C\left( {\alpha, \beta } \right) + {o_1}\left( 1 \right) - M{q^2}\sqrt[3]{\varepsilon }\\ = C\left( {\alpha, \beta } \right)\left( {1 -o\left( 1 \right)} \right)\left( {\varepsilon \to {o^ + }} \right).\end{array}$
2 主要结论定理2.1 ??设$\begin{array}{l}p > 1, \frac{1}{p} + \frac{1}{q} = 1, \alpha > 0, \beta >-1, \varphi \left( x \right) = {x^{\frac{{p\left( {\beta + 1} \right)}}{2}-1}}, \psi \left( y \right) = {y^{\frac{{q\left( {\beta + 1} \right)}}{2}-1}}, f, g > 0, f \in L_\varphi ^p\left( {0, \infty } \right), g \in L_\psi ^q\left( {0, \infty } \right)\end{array}$, 则有
$\begin{array}{l}\int_0^\infty {\int_0^\infty {\frac{{{{\left| {\ln \frac{x}{y}} \right|}^\alpha }{{\left( {\min \left\{ {x, y} \right\}} \right)}^\beta }f\left( x \right)g\left( y \right)}}{{x + y}}} } {\rm{d}}x{\rm{d}}y\\\;\;\;\;\;\;\;\;\;\;\;\;\; < C\left( {\alpha, \beta } \right){\left\| f \right\|_{p, \varphi }}{\left\| g \right\|_{q, \psi }}, \end{array}$ (10)
这里的常数因子C(α, β)(同式(7))是式(10)的最佳值.
证明 ??由H?lder不等式13和引理2及Fubini定理有
$\begin{array}{l}I: = \int_0^\infty {\int_0^\infty {\frac{{{{\left| {\ln \frac{x}{y}} \right|}^\alpha }{{\left( {\min \left\{ {x, y} \right\}} \right)}^\beta }f\left( x \right)g\left( y \right)}}{{x + y}}} } {\rm{d}}x{\rm{d}}y\\ = \int_0^\infty {\int_0^\infty {\frac{{{{\left| {\ln \frac{x}{y}} \right|}^\alpha }{{\left( {\min \left\{ {x, y} \right\}} \right)}^\beta }f\left( x \right)g\left( y \right)}}{{x + y}} \times } } \\\;\;\;\;\;\;\;\;\left[{\frac{{{y^{-\frac{{\beta + 1}}{{2p}}}}}}{{{x^{-\frac{{\beta + 1}}{{2q}}}}}}} \right]\left[{\frac{{{x^{-\frac{{\beta + 1}}{{2q}}}}}}{{{y^{-\frac{{\beta + 1}}{{2p}}}}}}} \right]{\rm{d}}x{\rm{d}}y\\ \le {\left[{\int_0^\infty {\int_0^\infty {\frac{{{{\left| {\ln \;\frac{x}{y}} \right|}^a}{{\left( {\min \;\left\{ {x, y} \right\}} \right)}^\beta }{f^p}\left( x \right)}}{{x + y}}\frac{{{y^{-\frac{{\beta + 1}}{2}}}{\rm{d}}x{\rm{d}}y}}{{{x^{-\frac{{p\left( {\beta + 1} \right)}}{{2q}}}}}}} } } \right]^{\frac{1}{p}}} \times \\{\left[{\int_0^\infty {\int_0^\infty {\frac{{{{\left| {\ln \frac{s}{y}} \right|}^\alpha }{{\left( {\min \left\{ {x, y} \right\}} \right)}^\beta }{g^q}\left( y \right)}}{{x + y}}\frac{{{x^{-\frac{{\beta + 1}}{2}}}}}{{{y^{-\frac{{q\left( {\beta + 1} \right)}}{{2p}}}}}}{\rm{d}}x{\rm{d}}y} } } \right]^{\frac{1}{q}}}\\ = {\left\{ {\int_0^\infty {\omega \left( {\alpha, \beta, x} \right){f^p}\left( x \right){\rm{d}}x} } \right\}^{\frac{1}{p}}} \times \\\;\;{\left\{ {\int_0^\infty {\omega \left( {\alpha, \beta, y} \right){g^q}\left( y \right){\rm{d}}y} } \right\}^{\frac{1}{q}}}\\ = C\left( {\alpha, \beta } \right){\left\| f \right\|_{p, \varphi }}{\left\| g \right\|_{q, \psi }}, \end{array}$ (11)
若式(11)取等号, 则存在不全为零的实数AB, 使$A\frac{{{y^{-\frac{{\beta + 1}}{2}}}}}{{{x^{-\frac{{p\left( {\beta + 1} \right)}}{{2q}}}}}}{f^p}\left( x \right) = B\frac{{{x^{-\frac{{\beta + 1}}{2}}}}}{{{y^{ - \frac{{q\left( {\beta + 1} \right)}}{{2p}}}}}}{g^q}\left( y \right), {\rm{a}}{\rm{.e}}$.于$\left( {0, \infty } \right) \times \left( {0, \infty } \right)$, 于是有常数C,使$A{x^{\frac{{p\left( {\beta + 1} \right)}}{2}}}{f^p}\left( x \right) = B{y^{\frac{{q\left( {\beta + 1} \right)}}{2}}}{g^q}\left( y \right) = C, {\rm{a}}{\rm{.e}}{\rm{.}}$.于$\left( {0, \infty } \right) \times \left( {0, \infty } \right)$, 不妨设A≠0, 则有${x^{\frac{{p\left( {\beta + 1} \right)}}{2}-1}}{f^p}\left( x \right) = \frac{C}{{Ax}}, {\rm{a}}{\rm{.e}}{\rm{.}}$$\left( {0, \infty } \right)$, 这与${\rm{0 < }}{\left\| f \right\|_{p, \varphi }} < \infty $矛盾, 故式(11)取严格不等号.若C(α, β)不是式(10)的最佳常数因子, 则存在正数K < C(α, β), 使式(10)的常数因子换成K后仍成立, 于是由式(8)和(9)有C(α, β)(1-o(1)) < K, 让ε→0+得: KC(α, β), 这与K < C(α, β)矛盾, 故C(α, β)是式(10)的最佳常数因子.
定理2.2 ??在与定理2.1相同的条件下, 我们还有
$\begin{array}{l}{\int_0^\infty y ^{\frac{{- \frac{{q\left( {\beta + 1} \right)}}{2} + 1}}{{q- 1}}}}{\rm{d}}y{\left[{\int_0^\infty {\frac{{{{\left| {\ln \;\frac{x}{y}} \right|}^\alpha }{{\left( {\min \;\left\{ {x, y} \right\}} \right)}^\beta }f\left( x \right)}}{{x + y}}{\rm{d}}x} } \right]^p}\\\;\;\;\;\; < {C^p}\left( {\alpha, \beta } \right)\left\| f \right\|_{p, \varphi }^p, \end{array}$ (12)
这里的常数因子Cp(α, β)是式(12)的最佳值,且式(12)与(10)等价.
证明 ??设置如下有界可测函数
${\left[{f\left( x \right)} \right]_n}:\min \left\{ {n, f\left( x \right)} \right\} = \left\{ \begin{array}{l}f\left( x \right), \;\;\;f\left( x \right) < n\\n, \;\;\;\;\;\;\;\;\;f\left( x \right) \ge n\end{array} \right..$
${\rm{0 < }}{\left\| f \right\|_{p, \varphi }} < \infty $, 存在n0∈N, 使得当nn0时, 有$0 < \int_{\frac{1}{n}}^n {{x^{\frac{{p\left( {\beta + 1} \right)}}{2}- 1}}\left[{f\left( x \right)} \right]_n^p{\rm{d}}x < \infty } $, 置${g_n}\left( y \right): = {y^{\frac{{- \frac{{q\left( {\beta + 1} \right)}}{2} + 1}}{{q- 1}}}}{\left[{\int_{\frac{1}{n}}^n {\frac{{{{\left| {\ln \;\frac{x}{y}} \right|}^\alpha }{{\left( {\min \;\left\{ {x, y} \right\}} \right)}^\beta }}}{{x + y}}{{\left[{f\left( x \right)} \right]}_n}{\rm{d}}x} } \right]^{\frac{p}{q}}}\left( {\frac{1}{n} < y < n, n \ge {n_0}} \right)$, 则当nn0时, 由式(10)有
$\begin{array}{l}0 < \int_{\frac{1}{n}}^n {{y^{\frac{{q\left( {\beta + 1} \right)}}{2}- 1}}g_n^q\left( y \right){\rm{d}}y} \\\;\;\;\;\; = \int_{\frac{1}{n}}^n {{y^{\frac{{- \frac{{q\left( {\beta + 1} \right)}}{2} + 1}}{{q- 1}}}}\left[{\int_{\frac{1}{n}}^n {\frac{{{{\left| {\ln \frac{x}{y}} \right|}^\alpha }{{\left( {\min \left\{ {x, y} \right\}} \right)}^\beta }}}{{x + y}} \times } } \right.} \\\;\;\;\;\;\;\;\;\;{\left. {\;{{\left[{f\left( x \right)} \right]}_n}{\rm{d}}x} \right]^p}{\rm{d}}y\\\;\;\;\; = \int_{\frac{1}{n}}^n {\int_{\frac{1}{n}}^n {\frac{{{{\left| {\ln \frac{x}{y}} \right|}^\alpha }{{\left( {\min \left\{ {x, y} \right\}} \right)}^\beta }}}{{x + y}}{{\left[{f\left( x \right)} \right]}_n}{g_n}\left( y \right){\rm{d}}x{\rm{d}}y} } \\\;\;\;\; < C\left( {\alpha, \beta } \right){\left\{ {\int_{\frac{1}{n}}^n {{x^{\frac{{p\left( {\beta + 1} \right)}}{2} - 1}}\left[{f\left( x \right)} \right]_n^p{\rm{d}}x} } \right\}^{\frac{1}{p}}} \times \\\;\;\;\;\;\;{\left\{ {\int_{\frac{1}{n}}^n {{y^{\frac{{q\left( {\beta + 1} \right)}}{2} -1}}g_n^q\left( y \right){\rm{d}}y} } \right\}^{\frac{1}{q}}}, \end{array}$ (13)
$\begin{array}{l}0 < \int_{\frac{1}{n}}^n {{y^{\frac{{q\left( {\beta + 1} \right)}}{2}-1}}g_n^q\left( y \right){\rm{d}}y} \\\;\;\; < {C^p}\left( {\alpha, \beta } \right)\left\| f \right\|_{p, \varphi }^p < \infty, \end{array}$ (14)
${\rm{0 < }}{\left\| f \right\|_{p, \varphi }} < \infty $.当 $n \to \infty $时, 应用式(10), 则式(13)取严格不等号, 式(14)亦然, 故有式(12).
反之, 由带权H?lder不等式有
$\begin{array}{l}I = \int_0^\infty {\int_0^\infty {\frac{{{{\left| {\ln \frac{x}{y}} \right|}^\alpha }{{\left( {\min \left\{ {x, y} \right\}} \right)}^\beta }}}{{x + y}}} } f\left( x \right)g\left( y \right){\rm{d}}x{\rm{d}}y\\\;\; = \int_0^\infty {\left[{{y^{\frac{{-\frac{{q\left( {\beta + 1} \right)}}{2} + 1}}{{p\left( {q-1} \right)}}}}\int_0^\infty {\frac{{{{\left| {\ln \frac{x}{y}} \right|}^\alpha }{{\left( {\min \left\{ {x, y} \right\}} \right)}^\beta }f\left( x \right){\rm{d}}x}}{{x + y}}} } \right]} \; \times \\\;\;\;\;\;\left[{{y^{\frac{{\frac{{q\left( {\beta + 1} \right)}}{2}-1}}{{p\left( {q-1} \right)}}}}g\left( y \right)} \right]{\rm{d}}y\\ \le \left\{ {\int_0^\infty {{y^{\frac{{ - \frac{{q\left( {\beta + 1} \right)}}{2} + 1}}{{\left( {q - 1} \right)}}}}{\rm{d}}y \times } } \right.\\{\left. {{{\left[{\int_0^\infty {\frac{{{{\left| {\ln \frac{x}{y}} \right|}^\alpha }{{\left( {\min \left\{ {x, y} \right\}} \right)}^\beta }}}{{x + y}}f\left( x \right)} {\rm{d}}x} \right]}^p}} \right\}^{\frac{1}{p}}}{\left\| g \right\|_{q, \psi }}\\ < C\left( {\alpha, \beta } \right){\left\| f \right\|_{p, \varphi }}{\left\| g \right\|_{q, \psi }}.\end{array}$
上不等式即为式(10), 因此式(10)和式(12)等价.
若式(12)中的常数因子不是最佳的, 则由式(12)得到式(10)的常数因子也不是最佳的, 故常数因子Cp(α, β)是式(12)的最佳值.
我们在式(10)和(12)中选取符合定理条件的参数α, β以及共轭指数对(p, q)的合适值, 并借助Maple数学软件的计算, 可以得到一些有意义的不等式.
如取α=1, β=0, p=q=2, 计算式(7)得$C\left( {1, 0} \right) = {c_0} = \frac{{{{\rm{\pi }}^2}}}{2} + 4{\rm{catalan}}-\frac{1}{2}\Psi \left( {1, \frac{3}{4}} \right) = 7.327\;724\;{76^ + }\left( {其中\Psi \left( {n, z} \right)为n次\Gamma 函数} \right)$, 则有式(2)和它的等价式:
$\int_0^\infty {{\rm{d}}y{{\left[{\int_0^\infty {\frac{{\left( {\ln \frac{x}{y}} \right)f\left( x \right)}}{{x + y}}{\rm{d}}x} } \right]}^2} < c_0^2\left\| f \right\|_2^2.} $ (15)
这里的常数因子c02是式(15)的最佳值.
如取α=2, β=1, p=q=2, 计算式(7)得$C\left( {2, 1} \right) = 3\zeta \left( 3 \right) = 3.606\;170\;{709^ + }$, 这时$\varphi \left( x \right) = x$, 设$f, g \in L_\varphi ^2\left( {0, \infty } \right), {\left\| f \right\|_{2, \varphi }}, {\left\| g \right\|_{2, \varphi }} > 0$, 则有下列等价式:
$\begin{array}{l}\int_0^\infty {\int_0^\infty {\frac{{{{\left( {\ln \frac{x}{y}} \right)}^2}\min \left\{ {x, y} \right\}}}{{x + y}}f\left( x \right)g\left( y \right){\rm{d}}x{\rm{d}}y} } \\\;\;\;\;\;\;\; < 3\zeta \left( 3 \right){\left\| f \right\|_{2, \varphi }}{\left\| g \right\|_{2, \varphi }}, \;\end{array}$ (16)
$\begin{array}{l}\int_0^\infty {{y^{- 1}}{\rm{d}}y{{\left[{\int_0^\infty {\frac{{{{\left( {\ln \frac{x}{y}} \right)}^2}\min \left\{ {x, y} \right\}}}{{x + y}}f\left( x \right){\rm{d}}x} } \right]}^2}} \\\;\;\;\;\;\;\;\;\;\; < 9{\zeta ^2}\left( 3 \right)\left\| f \right\|_{2, \varphi }^2.\;\end{array}$ (17)
这里的常数因子3ζ(3), 9ζ2(3)分别是式(16), (17)的最佳值.
如取$\alpha = \beta = \frac{1}{2}, p = q = 2$, 计算式(7)有$C\left( {\frac{1}{2}, \frac{1}{2}} \right) = \frac{{\sqrt {2{\rm{\pi }}} }}{4}\left[{\zeta \left( {\frac{3}{2}, \frac{3}{8}} \right)-\zeta \left( {\frac{3}{2}, \frac{7}{8}} \right)} \right] = 2.206\;556\;{861^ + }$, 这时φ(x)=x, 设$f, g \in L_\varphi ^2\left( {0, \infty } \right), {\left\| f \right\|_{2, \varphi }}, {\left\| g \right\|_{2, \varphi }} > 0$, 则有下列等价式:
$\begin{array}{l}\int_0^\infty {\int_0^\infty {\frac{{\sqrt {\left| {\ln \frac{x}{y}} \right|\min \left\{ {x, y} \right\}} }}{{x + y}}f\left( x \right)g\left( y \right){\rm{d}}x{\rm{d}}y} } \\ < \frac{{\sqrt {2{\rm{\pi }}} }}{4}\left[{\zeta \left( {\frac{3}{2}, \frac{3}{8}} \right)-\zeta \left( {\frac{3}{2}, \frac{7}{8}} \right)} \right]{\left\| f \right\|_{2, \varphi }}{\left\| g \right\|_{2, \varphi }}, \end{array}$ (18)
$\begin{array}{l}\int_0^\infty {\frac{1}{{\sqrt y }}{\rm{d}}y{{\left[{\int_0^\infty {\frac{{\sqrt {\left| {\ln \frac{x}{y}} \right|\min \left\{ {x, y} \right\}} }}{{x + y}}} f\left( x \right){\rm{d}}x} \right]}^2}} \\\;\; < \frac{{\rm{\pi }}}{8}{\left[{\zeta \left( {\frac{3}{2}, \frac{3}{8}} \right)-\zeta \left( {\frac{3}{2}, \frac{7}{8}} \right)} \right]^2}\left\| f \right\|_{2, \varphi }^2.\end{array}$ (19)
这里的常数因子$\frac{{\sqrt {2{\rm{\pi }}} }}{4}\left[{\zeta \left( {\frac{3}{2}, \frac{3}{8}} \right)-\zeta \left( {\frac{3}{2}, \frac{7}{8}} \right)} \right], \frac{{\rm{\pi }}}{8}{\left[{\zeta \left( {\frac{3}{2}, \frac{3}{8}} \right)-\zeta \left( {\frac{3}{2}, \frac{7}{8}} \right)} \right]^2}$分别是式(18), (19)的最佳值.
如取α=1, β=-12, p=q=2, 计算式(7)有$C\left( {1, \;\;-\;\;\frac{1}{2}} \right) = \frac{{\Psi \left( {1, \frac{1}{8}} \right)-\Psi \left( {1, \frac{5}{8}} \right)}}{2} = 30.993\;475\;{13^ + }$, 这时$\varphi \left( x \right) = \frac{1}{{\sqrt x }}$, 设$f, g \in L_\varphi ^2\left( {0, \infty } \right), {\left\| f \right\|_{2, \varphi }}, {\left\| g \right\|_{2, \varphi }} > 0$, 则有下列等价式:
$\begin{array}{l}\int_0^\infty {\int_0^\infty {\frac{{\left| {\ln \frac{x}{y}} \right|}}{{\left( {x + y} \right)\sqrt {\min \left\{ {x, y} \right\}} }}} f\left( x \right)g\left( y \right){\rm{d}}x{\rm{d}}y} \\ < \frac{{\Psi \left( {1, \frac{1}{8}} \right)-\Psi \left( {1, \frac{5}{8}} \right)}}{2}{\left\| f \right\|_{2, \varphi }}{\left\| g \right\|_{2, \varphi }}, \end{array}$ (20)
$\begin{array}{l}\int_0^\infty {\sqrt y {\rm{d}}y{{\left[{\int\limits_0^\infty {\frac{{\left| {\ln \frac{x}{y}} \right|}}{{\left( {x + y} \right)\sqrt {\min \left\{ {x, y} \right\}} }}f\left( x \right){\rm{d}}x} } \right]}^2}} \\\;\; < \frac{1}{4}{\left[{\Psi \left( {1, \frac{1}{8}} \right)-\Psi \left( {1, \frac{5}{8}} \right)} \right]^2}\left\| f \right\|_{2, \varphi }^2.\end{array}$ (21)
这里的常数因子$\frac{{\Psi \left( {1, \frac{1}{8}} \right)- \Psi \left( {1, \frac{5}{8}} \right)}}{2}, \frac{1}{4}{\left[{\Psi \left( {1, \frac{1}{8}} \right)-\Psi \left( {1, \frac{5}{8}} \right)} \right]^2}$分别是式(20), (21)的最佳值.
参考文献
[1] Weyl H. Singulare integral Gleichungenn Mit Besonderer Berucksichtigung Des Fourierschen integral theorems[M].Gottingen: Inaugeral-Dissertation, 1908.
[2] Hardy G H. Note on a theorem of Hilbert concerning series of positive term[J].Proc London Math Soc, 1925, 23:45–46.
[3] 杨必成. 一个具有混合核的Hilbert型积分不等式及其推广[J].四川师范大学学报:自然科学版, 2006, 31(3):281–284.
[4] 刘琼, 杨必成. 一个多参数混合核的Hilbert型积分不等式及其应用[J].浙江大学学报:理学版, 2012, 39(2):135–141.
[5] Liu Q, Chen D. A Hilbert-type intetral inequality with a hybrid kernel and its applications[J].Colloquium Mathematicum, 2016, 143(2):193–207.
[6] 杨必成. 参量化Hilbert型不等式研究综述[J].数学进展, 2009, 38(3):257–258.
[7] 杨必成. 关于一个非齐次核的Hilbert型积分不等式[J].上海大学学报:自然科学版, 2011, 17(5):603–606.
[8] 刘琼, 龙顺潮. 一个核为双曲正割函数的Hilbert型积分不等式[J].浙江大学学报:理学版, 2013, 40(3):255–259.
[9] 刘琼, 龙顺潮. 一个核为双曲余割函数的Hilbert型积分不等式[J].数学学报:中文版, 2013, 56(1):97–104.
[10] Liu Q. Two new integral inequalities and a relationship among operator norms[J].Journal of University of Chinese Academy of Sciences, 2015, 32(3):398–403.
[11] 苏变萍, 陈东立. 复变函数与积分变换[M].北京: 高等教育出版社, 2003.
[12] 匡继昌. 实分析引论[M].长沙: 湖南教育出版社, 1996.
[13] 匡继昌. 常用不等式[M].3版.济南: 山东科学技术出版社, 2004.


相关话题/理学 数学 湖南 浙江 不等式

  • 领限时大额优惠券,享本站正版考研考试资料!
    大额优惠券
    优惠券领取后72小时内有效,10万种最新考研考试考证类电子打印资料任你选。涵盖全国500余所院校考研专业课、200多种职业资格考试、1100多种经典教材,产品类型包含电子书、题库、全套资料以及视频,无论您是考研复习、考证刷题,还是考前冲刺等,不同类型的产品可满足您学习上的不同需求。 ...
    本站小编 Free壹佰分学习网 2022-09-19
  • 热力耦合问题数学均匀化方法的物理意义*
    复合材料具有比强度高、比刚度大等优点,广泛应用于航天、航空工业领域。众所周知,对于很多复合材料的宏观解,如低阶频率和模态,可以使用等应变模型或等应力模型[1]及其他均匀化方法[2]求解,但相对于宏观应力分析,细观结构分析要复杂很多。为了在计算精度和效率之间达到平衡,各种多尺度方法相继被提出,如数学均 ...
    本站小编 Free考研考试 2021-12-25
  • 民机横航向静稳定性适航符合性数学仿真评估*
    横航向静稳定性是评定飞机操稳特性的重要指标之一。横向静稳定性是指飞机在平衡状态受到外界非对称扰动产生小量滚转角Δ?后,具有自动恢复机翼水平姿态的趋势,判据为无量纲横向静稳定性导数Clβ<0。航向静稳定性是指飞机在平衡状态受到外界非对称扰动产生小量侧滑角Δβ后,具有自动消除侧滑运动的趋势,判据为无量纲 ...
    本站小编 Free考研考试 2021-12-25
  • 民机起飞爬升梯度适航符合性数学仿真评估*
    为了保障飞行安全,飞机起飞离地后需要迅速拉起爬升加速至起飞安全速度,达到10.5m的起飞安全高度,进入起飞爬升阶段并继续爬升至离地高度不低于450m。爬升梯度定义为飞机爬升的高度与飞过的水平距离的比值[1]。民机起飞阶段和起飞爬升阶段的爬升梯度反映了飞机超越地面障碍物并爬升到安全飞行高度的能力。根据 ...
    本站小编 Free考研考试 2021-12-25
  • 新型三轴离心机系统构型及数学建模
    现代军事、国防领域对某些无人高速飞行器的机动性能要求很高,即要求其具有很强的承受机动过载的能力[1,2].国内外的实践证明,如果某些产品只做地面普通试验,不测试其承受高过载下的性能,可能会导致产品在机动飞行中失效[3],为了在地面上验证无人高速飞行器的整体强度,就需要有一套可以模拟其在运动中承受载荷 ...
    本站小编 Free考研考试 2021-12-25
  • 精确华林不等式的一个推广
    doi:10.12202/j.0476-0301.2019276齐宗会1,汪晖2,刘永平3,,1.天津商业大学宝德学院,300384,天津2.天津师范大学数学科学学院,300387,天津3.北京师范大学数学科学学院,100875,北京基金项目:国家自然科学基金资助项目(11871006)详细信息通讯 ...
    本站小编 Free考研考试 2021-12-25
  • 基于力、能量、动量和作用量框架的物理学基础理论结构统一性探讨
    doi:10.12202/j.0476-0301.2019169曹盛林,北京师范大学天文系,100875,北京详细信息通讯作者:曹盛林(1937-),教授,学士.研究方向:相对论天体物理和宇宙学.e-mail:78108@bnu.edu.cn中图分类号:O41计量文章访问数:280HTML全文浏览量 ...
    本站小编 Free考研考试 2021-12-25
  • 中国数学学科成果评价方式研究
    doi:10.12202/j.0476-0301.2020048赵静1,刘姝2,,1.北京大学数学科学学院,100871,北京2.北京大学图书馆,100871,北京基金项目:北京大学科研管理项目“促进数学学科深远发展的科研机制研究”的资助项目(2016005)详细信息通讯作者:刘姝(1979-),女 ...
    本站小编 Free考研考试 2021-12-25
  • 一种复合营养素制剂的毒理学研究
    路子佳,谢瑶.一种复合营养素制剂的毒理学研究[J].北京理工大学学报(自然科学版),2019,39(S1):153~158.LUZi-jia,XIEYao.ToxicologicalStudyonthePreparationofCompoundNutrients[J].TransactionsofB ...
    本站小编 Free考研考试 2021-12-21
  • 姜黄灵芝中药保健食品毒理学研究
    路子佳,石萌萌,谢瑶.姜黄灵芝中药保健食品毒理学研究[J].北京理工大学学报(自然科学版),2017,37(s1):161~168.LUZi-jia,SHIMeng-meng,XIEYao.ToxicologlicalStudyontheChineseMedicineHealthFoodofCurc ...
    本站小编 Free考研考试 2021-12-21
  • 南宋《乐记》理学化阐释的两种路向——朱熹与杨简《乐记》中礼乐思想比较
    南宋《乐记》理学化阐释的两种路向——朱熹与杨简《乐记》中礼乐思想比较1中国人民大学艺术学院;2西北工业大学艺术教育中心出版日期:2020-11-16发布日期:2020-11-11作者简介:刘琉:艺术学博士,中国人民大学艺术学院副教授(北京100872);孙小迪(通讯作者):艺术学博士,西北工业大学艺 ...
    本站小编 Free考研考试 2021-12-21