数学物理方法 14 勒让德多项式

$\color{blue}{第十四章 勒让德多项式}$

$物理问题:$
$求一表面充电至点位为(1 + 3 \cos^2 \theta)的单位空心球内任一点的点位。$
$\left \lbrace \begin{array}{l}\Delta u = -, r < 1 \\ u |_{r = 1} = (1 + 3 \cos^2 \theta) \end{array} \right.$

$\Delta u = 0 \stackrel{令u = R(r)\Theta(\theta)\Phi(\varphi)}{----\to} u = ?$
$\dfrac{1}{\sin \theta} \dfrac{d}{d \theta}(\sin \theta \dfrac{d \Theta}{d \theta}) + [l(l+1) - \dfrac{m^2}{\sin^2 \theta}]\Theta = 0 \stackrel{m=0}{--\to}\Theta(\theta) = ?$
$令x = \cos \theta, y(x) = \Theta(\theta)$
$(1 - x^2)y^{\prime \prime} - 2xy^{\prime} + [l(l+1) - \dfrac{m^2}{1-x^2}]y = 0 \to y(x) = ?$
$当m=0时(1-x^2)y^{\prime \prime} - 2xy^{\prime} +l(l+1)y = 0 \to y(x) = ?$
$-勒让德方程$

$\color{blue}{\S 14.1 勒让德多项式-勒让德方程的解}$

$附:二阶线性常微分方程的技术解法1:$
$对于:W^{\prime \prime}(z) + p(z)W^{\prime}(z) + q(z)W(z) = 0 \quad (1)$
$若其系数p(z)和q(z)均在某点z_0及其邻域内解析,则称z_0为方程的常点.$
$在常点z=z_0的邻域|z-z_0| < R内,方程有唯一的一个满足初始条件$
$W(z_0) = C_0,W^{\prime}(z_0) = C_1形式为W(z) = \sum \limits_{k=0}^{\infty}C_k(z-z_0)^k \quad (2)$
$的幂级数解。其中C_0和C_1是任意常数;而其它各次幂系数于C_0和C_1的关系，$
$均由将形式解(2)代入方程(1)中通过比较方程两边同次幂的系数[即让左边$
$(z-z_0)的各次幂的系数均为零]来确定.$

$\color{blue}{\S 14.1.1 勒让德方程的级数解}$

$(1-x^2)y^{\prime \prime} - 2xy^{\prime} + l(l+1)y = 0 \quad (1)$
$p(x) = \dfrac{-2x}{1-x^2}, q(x) = \dfrac{l(l+1)}{1-x^2}, x = 0为该方程的常点$
$令y = \sum \limits _{k=0}^{\infty} c_k x^k \quad (2)$
$\sum \limits _{k=2}^{\infty}k(k-1)c_kx^{k-2} - \sum \limits _{k=2}^{\infty}k(k-1)c_kx^k - 2\sum \limits _{k=1}^{\infty} kc_kx^k + l(l+1)\sum \limits _{k=0}^{\infty} c_kx^k = 0$
$x^0:2\cdot1c_2 + l(l+1)c_0 = 0 \to c_2 = -\dfrac{l(l+1)}{2 \cdot 1} c_0$
$x^1:3\cdot 2 c_3 - 2c_1 + l(l+1)c_1 = 0 \to c_3 = -\dfrac{l(l+1) - 2}{3 \cdot 2}c_1$
$x^k: c_{k+2} = -\dfrac{[l(l+1) - k(k+1)}{(k+2)\cdot(k+1)}c_k \quad (3) - 系数递推公式$
$\therefore c_4 = -\dfrac{l^2+l-2 \cdot 3}{4 \cdot 3} c_2 = (-1)^2 \dfrac{(l-2)l(l+1)(l+3)}{4!}c_0$
$c_5 = (-1)^2\dfrac{(l-3)(l-1)(l+2)(l+4)}{5!}c_1$
$c_{2n}= (-1)^n\dfrac{(l-2n+2)(l-2n+4)\cdots l(l+1)(l+3)\cdots(l+2n-1)}{(2n)!}c_0 (4)$
$c_{2n+1}=(-1)^n\dfrac{(l-2n+1)(l-2n+3)\cdots(l-1)(l+2)(l+4)\cdots(l+2n)}{(2n+1)!}c_1 (5)$
$故y = \sum \limits_{k=0}^{\infty} c_kx^k = c_0 + \sum \limits _{n=1}^{\infty} c_{2n}x^{2n} + c_1x + \sum \limits_{n=1}^{\infty} c_{2n+1} x^{2n+1}$
$=y_0(x) + y_1(x), 式中:$
$y_0(x) = c_0 + \sum \limits_{n=1}^{\infty} c_{2n} x^{2n} \quad (6)$
$y_1(x) = c_1x + \sum \limits _{n=1}^{\infty} c_{2n+1} x^{2n+1} \quad (7)$

$c_{2n} = (-1)^n \dfrac{(l-2n+2)(l-2n+4)\cdots l(l+1)(l+3)\cdots(l+2n-1)}{(2n)!}c_0 \quad (4)$
$c_{2n+1} = (-1)^n\dfrac{(l-2n+1)(l-2n+3)\cdots(l-1)(l+2)(l+4)\cdots(l+2n)}{(2n+1)!}c_1 (5)$

$\color{blue}{14.1.2 解的敛散性}$

$1.由达氏判别法$
$R = \lim \limits _{k\to \infty}|\dfrac{a_k}{a_{k+1}}| = \lim \limits _{k \to \infty} | \dfrac{c_k}{c_{k+2}} | \stackrel{(3)}{=} \lim \limits _{k \to \infty} | \dfrac{(k+2)(k+1)}{l(l+1) -k(k+1)} | = 1$
$\to y(x) \left \lbrace \begin{array}{l}|x| < 1 \quad 收敛 \\ |x| > 1 \quad 发散 \\ |x| = 1 \quad 收敛?发散? \end{array} \right.$

$2.由高斯判别法$
$将x = \pm1 代入(6)和(7)得:$
$y_0(\pm 1) = c_0 + \sum \limits _{n=1}^{\infty} c_{2n}(\pm 1)^{2n} = c_0 \sum \limits _{n=0}^{\infty} f_n \quad f_n一常数$
$类似y_1(\pm 1) = c_1 \sum \limits _{n=1}^{\infty} g_n$
$\therefore \dfrac{f_n}{f_{n+1}} = \dfrac{(2n+2)(2n+1)}{2n(2n+1) -l(l+1)} = 1 + \dfrac{1}{n} + \dfrac{l(l+1)}{4n^2} \cdots = 1 + \dfrac{1}{n} + o(\dfrac{1}{n^2})$
$\dfrac{g_n}{g_{n+1}} = 1 + \dfrac{1}{n} + o(\dfrac{1}{n^2}) \quad \therefore y_0(x), y_1(x)在x = \pm 1发散$

$\color{blue}{14.1.3 本征值问题}$

$\left \lbrace \begin{array}{l}(1-x^2)y^{\prime \prime} - 2xy^{\prime} + l(l+1)y = 0, l(l+1) -常数 (8) \\ y |_{x=\pm 1} \to 有限 \end{array} \right.$
$c_{k+2} = \dfrac{[l(l+1) - k(k+1)]}{(k+2)\cdot(k+1)}c_k \quad (3)$
$若取l = 0, 1, 2, \dots,则当l = k时,$
$c_{k+2} = c_{l+2} = 0 \cdot c_l = 0$
$从而有c_{k+4} = 0, c_{k+6} = 0, \cdots$
$即y_0 或 y_1 \to l次多项式$

$1.若l = k = 2n, n = 0, 1, 2, \cdots$
$则c_{k+2} = c_{l+2} = c_{2n+2} = 0, \therefore c_{2n+4} = c_{2n+6} = \cdots = 0$
$y_0(x) = c_0 + c_2x^2 + \cdots + c_{2n}x^{2n}$
$=c_0 + c_2x^2 + \cdots + c_lx^l \to l次多项式$
$y_1(x) = c_1x + c_3x^3 + \cdots + c_{2n+1}x^{2n+1} + c_{2n+3}x^{2n+3} + \cdots \to 无穷级数$

$2.若l = k = 2n+1, n = 0, 1, 2, \cdots$
$则c_{k+2} = c_{l+2} = c_{2n+3} = 0, \therefore c_{2n+5} = c_{2n+7} = \cdots = 0$
$y_1(x) = c_1x + c_3x^3 + \cdots + c_{2n+1}x^{2n+1}$
$=c_1x + c_3x^3 + \cdots + c_lx^l \to l次多项式$
$y_0(x) = c_0 + c_2x^2 + \cdots + c_{2n+2}x^{2n+2} + c_{2n+4}x^{2n+4} + \cdots \to 无穷级数$
总之,本征值问题
$\left \lbrace \begin{array}{l}(1-x^2)y^{\prime \prime} - 2xy^{\prime} + l(l+1)y = 0, l(l+1) -常数 (8) \\ y |_{x=\pm 1} \to 有限 \end{array} \right.$
$本征值:l(l+1), l = 0, 1, 2, \cdots$
$本征函数:\left \lbrace \begin{array}{l}y_0(x) = c_0 + c_2x^2 + \cdots + c_lx^l, l = 2n \\ y_1(x) = c_1x + c_3x^3 + \cdots + c_lx^l, l = 2n+1 \end{array} \right.$
$c_{2n} = (-1)^n \dfrac{(l-2n+2)(l-2n+4) \cdots l(l+1)(l+3)\cdots(l+2n-1)}{(2n)!}c_0$
$c_{2n+1} = (-1)^n\dfrac{(l-2n+1)(l-2n+3)\cdots(l-1)(l+2)(l+4)\cdots(l+2n)}{(2n+1)!}c_1$

$\color{blue}{14.1.4 勒让德多项式}$

$选c_l = \dfrac{(2l)!}{2^l(l!)^2},记上述l次多项式y_0(x)或y_1(x)为P_l(x),称之为l阶Legendre多项式.$
$则由(3):c_k = \dfrac{(k+2)(k+1)}{k(k+1) - l(l+1)}c_{k+2} = \dfrac{(k+2)(k+1)}{(k-l)(k+l+1)}c_{k+2}$
$\to c_{l-2} = \dfrac{l(l-1)}{-2(2l-1)}c_l = (-1)\dfrac{(2l-2)!}{2^l(l-1)!(l-2)!},$
$c_{l-4} = \dfrac{(l-2)(l-3)}{-4(2l-3)}c_{l-2} = (-1)^2\dfrac{(2l-4)!}{2^l2\cdot(l-2)!(l-4)!},$
$c_{l-2n} = (-1)^n\dfrac{(2l-2n)!}{2^ln!(l-n)!(l-2n)!} \to p_l(x) = \sum \limits _n c_{l-2n}x^{l-2n}$
$P_l(x) = \sum \limits _{n=0}^{\frac{l}{2}} \dfrac{(-1)^n(2l-2n)!}{2^ln!(l-n)!(l-2n)!}x^{l-2n} \quad (9) -勒让德多项式$
$其中[\dfrac{l}{2}] = \left \lbrace \begin{array}{l}\dfrac{l}{2}, l = 2n \\ \dfrac{l-1}{2}, l = 2n+1 \end{array} \right.$
$于是:\left \lbrace \begin{array}{l}(1-x^2)y^{\prime \prime} - 2xy^{\prime} + l(l+1) y = 0, l(l+1) - 常数 (8) \\ y|_{x = \pm 1} \to 有限 \end{array} \right.$
$\Rightarrow \left \lbrace \begin{array}{l}本征值:l(l+1),l = 0, 1, \cdots \\ 本征函数:P_l(x) \end{array} \right.$
$P_0(x) = \dfrac{(-1)^0 0!}{2^0 0!0!0!}x^{0-0} = 1 \quad (10)$
$l = 1, n:0 \to \dfrac{1-1}{2} = 0, P_1(x) = \dfrac{(-1)^0(2-0)!}{2^1 0! (1-0)! (1-0)!} x^1 = x \quad (11)$
$l = 2, n: 0 \to \dfrac{2}{2} = 1,$
$P_2(x) = \dfrac{(-1)^0(4-0)!}{2^2 0! (2-0)! (2-0)!} x^2 + \dfrac{(-1)^1(4-2)!}{2^2 1! (2-1)! (2-1)!} x^0 = \dfrac{1}{2}(3x^2 - 1) \quad (12)$

$\color{blue}{P_l(1) \equiv 1}$

$\color{blue}{14.1.5 勒让德多项式的其它表示}$

$1.微分式$
$P_l(x) = \dfrac{1}{2^l l!} \dfrac{d^l}{d x^l}(x^2 -1)^l \quad (13)$

$2.积分式$
$P_l(x) = \dfrac{1}{2 \pi i}\oint_{l^{*}} \dfrac{(\xi^2 - 1)^l}{2^l (\xi - x)^{l+1}} d \xi \quad (14)$

$\color{blue}{14.1.6 小结}$

$\left \lbrace \begin{array}{l}(1-x^2)y^{\prime \prime} - 2xy^{\prime} + l(l+1)y = 0, l(l+1) -常数(8) \\ y|_{x = \pm 1} \to 有限 \end{array} \right.$
$\left \lbrace \begin{array}{l}本征值: l(l+1), l = 0, 1, \cdots \\ 本征函数:P_l(x) \end{array} \right.$
$P_l(x) = \sum \limits _{n=0}^{\frac{l}{2}} \dfrac{(-1)^n(2l-2n)!}{2^l n! (l-n)! (l-2n)!} x^{l-2n} \quad (9)$
$P_0(x) = 1, P_1(x) = x, P_2(x) = \dfrac{1}{2}(3x^2 - 1), P_l(1) \equiv 1$

$\color{blue}{\S 14.2 勒让德多项式的性质}$

$\color{blue}{14.2.1 母函数关系}$

$\dfrac{1}{\sqrt{1 - 2xt + t^2}} = \sum \limits _{l=0}^{\infty} P_l(x) t^l, |t| < 1 (1)$
$注:若w(x, t) = \sum \limits _{n} F_n(x) t^n,则称w(x, t) 为F_n(x)的母函数.$
$物理背景:设在单位球北极置有电量为4 \pi \omega_0的正电荷,则在r < 1内,任一点的点位u为:$
$\Delta u = 0, r < 1, 令u(r, \theta) = R(r)\Theta(\theta)$
$\to \left \lbrace \begin{array}{l}r^2R^{\prime \prime} + 2rR^{\prime} - l(l+1)R = 0 \\ (1-x^2)y^{\prime \prime} -2xy^{\prime} + l(l+1)y = 0, [x = \cos \theta, y(x) = \Theta(\theta)] \end{array} \right.$
$r^2R^{\prime \prime} +2rR^{\prime} -l(l+1)R = 0 \to R_l(r) = c_lr^l + d_l r^{-(l+1)}$
$(1-x^2)y^{\prime \prime} -2xy^{\prime} + l(l+1)y = 0 \to y(x) = P_l(x)$
$\therefore r < 1: u(r, \theta) = R(r)\Theta(\theta) = \sum \limits _{l=0}^{\infty} c_l r^l P_l(x)$
$又:u = \dfrac{1}{d} = \dfrac{1}{\sqrt{1 - 2r \cos \theta + r^2}}$
$\therefore \dfrac{1}{\sqrt{1 - 2rx + r^2}} = \sum \limits _{l=0}^{\infty} c_lr^lP_l(x)$
$取x = 1: \dfrac{1}{\sqrt{1 - 2r + r^2}} = \sum \limits _{l=0}^{\infty} c_l r^l \to \dfrac{1}{1-r} = \sum \limits_ {l=0}^{\infty} c_l r^l$
$\because r < 1, \dfrac{1}{1 - r} = \sum \limits _{l=0}^{\infty} r^l \to \sum \limits _{l=0}^{\infty} r^l = \sum \limits_{l=0}^{\infty}c_lr^l \to c_l \equiv 1, l = 0, 1, \cdots$
$于是:\dfrac{1}{\sqrt{1 - 2rx + r^2}} = \sum \limits _{l=0}^{\infty} P_l(x)r^l, (r < 1)$
$更一般:\dfrac{1}{\sqrt{1 - 2xt + t^2}} = \sum \limits _{l=0}^{\infty} P_l(x) t^l, |t| < 1$

$\color{blue}{14.2.2 递推公式}$

$\dfrac{1}{\sqrt{1 - 2xt + t^2}} = \sum \limits _{l=0}^{\infty} P_l(x) t^l, |t| < 1 \quad (1)$
$1. (l+1)P_{l+1}(x) - (2l+1)xP_l(x) + lP_{l-1}(x) = 0 \quad (2)$
$2.(2l+1)P_l(x) = P_{l+1}^{\prime}(x) - P_{l-1}^{\prime}(x) \quad (3)$
$\dfrac{d}{dt}(1) \to \dfrac{x-t}{(1-2xt+t^2)^{\frac{3}{2}}} = \sum \limits _{l=1}^{\infty} P_l(x)lt^{l-1}$
$两边 \times (1-2xt+t^2) \to$
$(x-t)\sum \limits _{l=0}^{\infty} P_l(x)t^l = (1-2xt+t^2)\sum \limits _{l=1}^{\infty}l P_l(x)t^{l-1}$
$x\sum \limits_{l=0}^{\infty}P_l(x)t^l - \sum \limits_{l=0}^{\infty} P_l(x)t^{l+1} = \sum \limits_{l=1}^{\infty}l P_l(x)t^{l-1} - 2x\sum \limits_{l=1}^{\infty}lP_l(x)t^l + \sum \limits_{l=1}^{\infty}l P_l(x)t^{l+1}$
$比较等式两边t^l的系数即得(2)式$

$\dfrac{d}{dx}(1) \to \dfrac{t}{(1-2xt+t^2)^{\frac{3}{2}}} = \sum \limits _{l=0}^{\infty}P_l^{\prime}(x)t^l$
$t \sum \limits_{l=0}^{\infty}P_l(x)t^l = (1-2xt+t^2)\sum \limits_{l=0}^{\infty}P_l^{\prime}(x)t^l$
$t^{l+1}:P_l(x) = P_{l+1}^{\prime}(x) - 2xP_l^{\prime}(x) + P_{l-1}^{\prime}(x) \quad (4)$
$\dfrac{d}{dx}(2) \to (l+1)P_{l+1}^{\prime}(x) - (2l+1)P_l(x) - (2l+1)xP_l^{\prime}(x) + lP_{l-1}^{\prime}(x) = 0$
$\to xP_l^{\prime}(x) = P_l(x) - \dfrac{l+1}{2l+1}P_{l+1}^{\prime}(x) - \dfrac{l}{2l+1}P_{l-1}^{\prime}(x) \quad (5)$
$(5)代入(4) \to (3)$

$用途(1)可用低阶的勒让德多项式求高阶的勒让德多项式之值.$
$如:P_0(x) = 1, P_1(x) = x \stackrel{(2)}{\to}P_2(x)$
$(2)可计算含P_l(x)的积分.$
$如:\int_a^b P_l(x) dx = \dfrac{1}{2l+1}\int_a^b[P_{l+1}^{\prime}(x) - P_{l-1}^{\prime}(x)]dx$
$=\dfrac{1}{2l+1}[P_{l+1}(x) - P_{l-1}(x)] _a^b$

$\color{blue}{14.2.3 正交性}$

$\int_{-1}^{1}P_l(x)P_k(x)dx = \dfrac{2}{2l+1}\delta_{kl}, k, l = 0, 1, 2, \cdots (6)$
$其中\delta_{kl} = \left \lbrace \begin{array}{l}0 \quad k \neq l \\ 1 \quad k = l \end{array} \right.(克罗内克尔函数)$
$\because \dfrac{d}{dx}[(1-x^2)P_l^{\prime}(x)] + l(l+1)P_l(x) = 0 \quad (7)$
$\dfrac{d}{dx}[(1-x^2)P_k^{\prime}(x)] + k(k+1)P_k(x) = 0 \quad (8)$
$\int_{-1}^{1}[(7) \cdot P_k(x) - (8)\cdot P_l(x)] dx :$
$[k(k+1) - l(l+1)]\int_{-1}^{1}P_k(x)P_l(x) dx = \int_{-1}^{1} \dfrac{d}{dx}[(1-x^2)P_l^{\prime}(x)]P_k(x)dx - \int_{-1}^{1} \dfrac{d}{dx}[(1-x^2)P_k^{\prime}(x)]P_l(x)dx = 0$

$\int_{-1}^{1}P_l(x)P_k(x)dx = \dfrac{2}{2l+1}\delta_{kl}, k, l = 0, 1, 2, \cdots (6)$
$\therefore 当k \neq l, \int_{-1}^{1}P_l(x)P_k(x) dx = 0$
$当k = l, \dfrac{1}{(\sqrt{1-2xt+t^2})^2} = \sum \limits _{l=0}^{\infty}P_l(x)t^l \cdot \sum \limits_{k=0}^{\infty}P_k(x)t^k$
$\int_{-1}^{1}\dfrac{1}{1-2xt+t^2}dx = \sum \limits_{l=0}^{\infty} \sum \limits_{k=0}^{\infty} \int_{-1}^{1}P_l(x)P_k(x)dxt^{l+k}$
$=\sum \limits_{l=0}^{\infty}\int_{-1}^{1}P_l^2(x)dxt^{2l}$

$\int_{-1}^{1}P_l(x)P_k(x)dx = \dfrac{2}{2l+1}\delta_{kl}, k, l = 0, 1, 2, \cdots (6)$
$左边 = \int_{-1}^{1} \dfrac{d(1-2xt+t^2)}{1-2xt+t^2} \cdot \dfrac{1}{-2t}$
$=\dfrac{1}{-2t} \ln(1-2xt+t^2)|_{-1}^{1} = \dfrac{1}{2t} \ln \dfrac{(1+t)^2}{(1-t)^2} = \dfrac{1}{t} \ln \dfrac{1+t}{1-t}$
$\stackrel{T展}{=}\sum \limits_{l=0}^{\infty}\dfrac{2}{2l+1}t^{2l}$
$当k =l:\int_{-1}^{1}P_l^2(x)dx = \dfrac{2}{2l+1}$
$右边= \sum \limits_{l=0}^{\infty} \int_{-1}^{1} P_l^2(x) dx t^{2l}$
$其中N_l = \sqrt{\dfrac{2}{2l+1}} \to P_l(x)模,\dfrac{1}{N_l} \to 归一化因子$
$\therefore 若记P_L(x) = \dfrac{1}{N_l}P_l(x), P_K(x) = \dfrac{1}{N_k}P_k(x)$
\int_{-1}^{1} P_L(x)P_K(x) dx = \delta_{KL} -正交归一
$用途：可计算含P_l(x)的积分.$
$\int_{-1}^{1}P_{199}(x)P_{300}(x)dx = 0$
$\int_{-1}^{1}P_8^2(x) dx = \dfrac{2}{2 \cdot 8 + 1} = \dfrac{2}{17}$
$\int_{-1}^{1}P_8(x)P_9(x) dx = 0$
$\int_{-1}^{1}xP_8(x)P_9(x)dx = \dfrac{9}{17} \cdot \dfrac{2}{18 + 1}$

$\color{blue}{14.2.4 广义傅氏展开}$

$f(x) = \sum \limits_{l=0}^{\infty} C_lP_l(x) \quad (9)$
$C_l = \dfrac{2l+1}{2}\int_{-1}^{1}f(x)P_l(x)dx \quad (10)$
$用途：(1)在物理中常需将作为表征的物理量展开为级数进行分析。$
$(2)在求解数学物理方程时其解常是某函数的无穷级数,如稳恒电场的解,\\ 就是Legendre级数.$
$例：求一表面充电至点位为(1+3\cos^2\theta)的单位空心球内任一点的点位。$
$解:\left \lbrace \begin{array}{l}\Delta u = 0, r < 1 \quad (1) \\ u|_{r=1} = (1+3\cos^2\theta) \quad (2) \end{array} \right. 令u(r, \theta) = R(r)\Theta(\theta)$
$(1) \to \left \lbrace \begin{array}{l}r^2R^{\prime \prime} + 2rR^{\prime} -l(l+1)R = 0 \to R_l(r) = c_lr^l + d_lr^{-(l+1)} \\ (1-x^2)y^{\prime \prime} - 2xy^{\prime} + l(l+1)y = 0 \to y(x) = P_l(x) \end{array} \right.$
$(1) \to u(r, \theta) = \sum \limits_{l=0}^{\infty}C_lr^lP_l(\cos \theta) \to \sum \limits_{l=0}^{\infty}C_lP_l(\cos \theta) = (1+3\cos^2 \theta)$
$即\sum \limits_{l=0}^{\infty}C_lP_l(x) = (1+3x^2) \quad (3)$
$C_l = \dfrac{2l+1}{2}\int_{-1}^{1}(1+3x^2)P_l(x)dx$
$1 = P_0(x), P_2(x) = \dfrac{1}{2}(3x^2-1) \to 3x^2 = 2P_2 +P_0$
$\therefore C_l = \dfrac{(2l+1)}{2}[\int_{-1}^{1}P_0P_l(x)dx + 2\int_{-1}^{1}P_2P_l(x)dx + \int_{-1}^{1}P_0P_l(x)dx]$
$=(2l+1)[\int_{-1}^{1}P_2P_l(x)dx + \int_{-1}^{1}P_0P_l(x)dx]$
$C_0 = (2 \cdot 0 +1)\int_{-1}^{1}P_0^2(x)dx = 2, C_2 = (2 \cdot 2 +1)\int_{-1}^{1}P_2^2(x)dx = 2$
$C_l \equiv 0(l \neq 0, 2) \to \color{blue}{u(r, \theta) = 2r^0 P_0(\cos \theta) + 2r^2P_2(\cos \theta)}$

$另一方法:1 = P_0(x), 3x^2 = 2P_2 + P_0$
$(3) \to \sum \limits_{l=0}^{\infty} C_lP_l(x) = 1 + 3x^2 = P_0(x) + 2P_2(x) + P_0(x) = 2P_0 + 2P_2$
$\therefore C_0 = 2, C_2 = 2, C_l \equiv 0(l \neq 0, 2)$
$从而有:u(r, \theta) = 2r^{0}P_0(\cos \theta) + 2r^2 P_2(\cos \theta)$
$= 2 + 2 r^2 \dfrac{1}{2}(3\cos^2 \theta - 1)$
$=2 + 3 r^2 \cos^2 \theta - r^2$

$\color{blue}{14.2.5 小结}$

$1.母函数关系:\dfrac{1}{\sqrt{1-2xt+t^2}} = \sum \limits_{l=0}^{\infty}P_l(x)t^l, |t| < 1 (1)$
$2.递推公式$
$1.(l+1)P_{l+1}(x) - (2l+1)xP_l(x) + lP_{l-1}(x) = 0 \quad(2)$
$2.(2l+1)P_l(x) = P_{l+1}^{\prime}(x) - P_{l+1}^{\prime}(x) \quad (3)$
$3.正交性$
$\int_{-1}^{1}P_l(x)P_k(x)dx = \dfrac{2}{2l+1}\delta_{kl}, k, l = 0, 1, 2, \cdots (6)$
$4.广义傅氏展开$
$f(x) = \sum \limits_{l=0}^{\infty} C_lP_l(x) \quad (9)$
$C_l = \dfrac{2l+1}{2}\int_{-1}^{1}f(x)P_l(x)dx \quad (10)$

$\color{blue}{\S 14.3 球函数 - 缔合勒让德方程的解}$

$问题引入：$
$由分离变量法有:$
$\Delta u = 0 \stackrel{令u=R(r)\Theta(\theta)\Phi(\varphi)}{----\to}$
$r^2R^{\prime \prime} + 2rR^{\prime} - l(l+1)R = 0 \to R(r) = c_lr^l + d_lr^{-(l+1)}$
$\Phi^{\prime \prime} + m^2 \Phi = 0 \to \Phi_m(\varphi) = A_m \cos m \varphi + B_m \sin m \varphi$
$\dfrac{1}{\sin \theta}\dfrac{d}{d \theta}(\sin \theta \dfrac{d \Theta}{d \theta}) + [l(l+1) - \dfrac{m^2}{\sin^2 \theta}]\Theta = 0 \to \Theta(\theta) = ?$
$令x =\cos \theta, y(x) = \Theta(\theta)$
$(1-x^2)y^{\prime \prime} - 2xy^{\prime} + [l(l+1) - \dfrac{m^2}{1-x^2}]y = 0 \to y(x) = ?$
$物理背景:半径为a的均匀球，表面温度保持u_0\sin^2 \theta \cos 2\varphi, u_0为常数，\\ 求球内稳定温度分布。$
$\left \lbrace \begin{array}{l}\Delta u(r, \theta, \varphi) = 0, r < a \quad (1) \\ u|_{r=a} = u_0 \sin^2 \theta \cos 2 \varphi \quad (2) \end{array} \right.$
$u(r, \theta, \varphi) = ?, r< a$

$\color{blue}{14.3.1 缔合勒让德方程的本征值问题}$

$1.缔合勒让德函数$
$\left \lbrace \begin{array}{l}(1-x^2)y^{\prime \prime} - 2xy^{\prime} + [l(l+1) -\dfrac{m^2}{1-x^2}]y = 0 \quad (1) \\ y|_{x = \pm 1} \to 有限, x =0-常点 \end{array} \right.$
$令y(x) = (1-x^2)^{\frac{m}{2}}v(x) \quad (2), 代入(1)得:$
$(1-x^2)v^{\prime \prime}(x) - 2(m+1)xv^{\prime}(x) + [l(l+1) - m(m+1)]v = 0 \quad (3)$
$又(1-x^2)P_l^{\prime \prime}(x) -2xP_l^{\prime}(x) + l(l+1)P_l(x) = 0 \quad (4)$
$\dfrac{d^m}{dx^m}(4): (1-x^2)[P_l^{(m)}(x)]^{\prime \prime} - 2(m+1)x[P_l^{(m)}(x)]^{\prime} +[l(l+1)-m(m+1)]P_l^{(m)}(x) = 0 \quad (5)$
$对比(3)(5)式:v(x) = P_l^{(m)}(x)$
$记P_l^m(x) = (1-x^2)^{\frac{m}{2}}P_l^{(m)}(x) - 称为缔合Legendre函数$
$本征值：l(l+1), l = 0, 1, 2, \cdots$
$本征函数:y(x) = P_l^{(m)}(x) = (1-x^2)^{\frac{m}{2}}\dfrac{d^m}{dx^m}P_l(x), m = 0, 1, \cdots, l(6)$
$P_l^0(x) = P_l(x)$
$P_1^1(x) = (1-x^2)^{\frac{1}{2}}\dfrac{d}{dx}P_1(x) = (1-x^2)^{\frac{1}{2}}, or \;P_1^1(\cos \theta) = \sin \theta$
$P_2^1(x) = (1-x^2)^{\frac{1}{2}}\dfrac{d}{dx}P_2(x) = 3(1-x^2)^{\frac{1}{2}}x, or \; P_2^1(\cos \theta) = \dfrac{3}{2}\sin 2\theta$
$P_2^2(x) = (1-x^2)\dfrac{d^2}{dx^2}P_2(x) = 3(1-x^2), or \; P_2^2(\cos \theta) = 3\sin^2 \theta$
$\color{blue}{P_0(x) = 1, P_1(x) = x, P_2(x) = \dfrac{1}{2}(3x^2-1), P_l(1) \equiv 1}$

$2.缔合勒让德函数P_l^m(x)的微分式$
$P_l^m(x) = \dfrac{(1-x^2)^{\frac{m}{2}}}{2^l l!} \dfrac{d^{l+m}}{dx^{l+m}}(x^2-1)^l \quad (7)$
$P_l^{-m}(x) = \dfrac{(1-x^2)^{-\frac{m}{2}}}{2^l l!} \dfrac{d^{l-m}}{dx^{l-m}}(x^2-1)^l$
$P_l^{-m}(x) = (-1)^m \dfrac{(l-m)!}{(l+m)!} P_l^m(x)$

$3.缔合勒让德函数P_l^m(x)的积分式$
$P_l^m(x) = \dfrac{(1-x^2)^{-\frac{m}{2}}}{2^l l!} \dfrac{(l+m)!}{2 \pi i} \oint_{l^{*}}\dfrac{(\xi^2 - 1)^l}{(\xi - x)^{l+m+1}} d \xi \quad (8)$

$\color{blue}{14.3.2 缔合勒让德函数的性质}$

$1.母函数关系式:\dfrac{(2m-1)!!}{(1-2xt+t^2)^{\frac{m+1}{2}}} = \sum \limits_{l=m}^{\infty} P_l^m(x) t^{l-m}$
$2.递推公式:$
$(l+1-m)P_{l+1}^m(x) - (2l+1)xP_l^m(x) + (l+m)P_{l-1}^{m}(x) = 0 \quad (9)$
$\because (l+1)P_{l+1}(x) - (2l+1)xP_l(x) + lP_{l-1}(x) = 0 \quad (A)$
$\dfrac{d^m}{dx^m}(A): (l+1)P_{l+1}^{(m)}(x) - (2l+1)xP_l^{(m)}(x) - m(2l+1)P_l^{(m-1)}(x) + lP_{l-1}^{(m)}(x) = 0 (10)$
$又(2l+1)P_l(x) = P_{l+1}^{\prime}(x) - P_{l-1}^{\prime}(x)$
$\therefore -m(2l+1)P_l^{(m-1)}(x) = -mP_{l+1}^{(m)}(x) + mP_{l-1}^{(m)}(x) \quad (11)$
$[(11)代入(10)]\cdot(1-x^2)^{\frac{m}{2}} \Rightarrow (9)$

$3.正交性$
$\int_{-1}^{1}P_l^m(x)P_k^m(x)dx = \dfrac{(l+m)!}{(l-m)!}\dfrac{2}{2l+1} \delta_{kl}$

$4.广义傅氏展开$
$f(x) = \sum \limits_{l=0}^{\infty} \sum \limits_{m=0}^{l} C_l^m P_l^m(x),$
$C_l^m = \dfrac{(l-m)!}{(l+m)!} \dfrac{2l+1}{2}\int_{-1}^{1} f(x) P_l^m(x)dx$

$5.例题:半径为a的均匀球，表面温度保持u_0 \sin^2 \theta \cos 2\varphi, u_0-常数,\\ 求球内稳定温度分布。$
$解:\left \lbrace \begin{array}{l}\Delta u(r, \theta, \varphi) = 0, r < 1 \quad (1) \\ u|_{r=a} = u_0 \sin^2 \theta \cos 2 \varphi \quad (2) \end{array} \right.$
$u = \sum \limits _{l=0}^{\infty} \sum \limits_{m=0}^{l}(A_l^m \cos m \varphi + B_l^m \sin m \varphi)r^l P_l^m(\cos \theta)$
$\to \sum \limits_{l=0}^{\infty} \sum \limits_{m=0}^{l}(A_l^m a^l \cos m \varphi + B_l^m a^l \sin m \varphi)P_l^m(\cos \theta)$
$=u_0 \sin^2 \theta \cos 2 \varphi = \dfrac{1}{3} u_0 P_2^2(\cos \theta) \cos 2 \varphi$
$\therefore A_2^2 a^2 = \dfrac{1}{3}u_0, A_l^m a^l = 0(m, l \neq 2), B_l^m a^l \equiv 0$
$即A_2^2 = \dfrac{u_0}{3a^2}, A_l^m = 0(l,m \neq 2), B_l^m \equiv 0$
$u(r, \theta, \varphi) = \dfrac{u_0r^2}{a^2} \sin^2 \theta \cos 2 \varphi$

$\color{blue}{14.3.3 球函数}$

$1.定义:$
$\Delta u = 0 \stackrel{令u = R(r)\Theta(\theta)\Phi(\varphi)}{---\to}$
$u_{l,m} = (c_lr^l + d_lr^{-(l+1)})(A_m\cos m\varphi +B_m \sin m \varphi)P_l^m(\cos \theta)$
$记y_l^m(\theta, \varphi) = P_l^m(\cos \theta) \left \lbrace \begin{array}{l}\sin m \varphi \\ \cos m \varphi \end{array} \right \rbrace, m = 0, 1, \cdots, l; l = 0, 1, \cdots$
$或y_l^m(\theta, \varphi) = P_l^{|m|}(\cos \theta)e^{im \varphi}, m = 0, \pm1, \cdots, \pm l -l阶球函数$
$(独立的y_l^m(\theta, \varphi)共有2l+1个)$
$u = \sum \limits_{l=0}^{\infty} \sum \limits _{m=-l}^l (c_lr^l + d_l r^{-(l+1)})y_l^m(\theta, \varphi)$

$2.性质$
$(1)正交归一性:定义归一化的球函数为:$
$y_{l,m}(\theta, \varphi) = \sqrt{\dfrac{2l+1}{4 \pi} \dfrac{(l-|m|)!}{(l+|m|)!}} P_l^{|m|} (\cos \theta) e^{im\varphi}$
$m = 0, \pm 1, \pm 2, \cdots, \pm l; l = 0, 1, \cdots$

$易于证明:$
$\int_0^{\pi} \int_0^{2\pi}y_{l,m}(\theta, \varphi) \overline{y_{k,n}(\theta, \varphi)} \sin \theta d \varphi d \theta = \delta_{kl} \delta_{nm}$
$y_{l, -m}(\theta, \varphi) = (-1)^m \overline{y_{l,m}(\theta, \varphi)}$
$y_{0,0} = \dfrac{1}{\sqrt{4\pi}}, y_{1,\pm 1} = \mp \sqrt{\dfrac{3}{8\pi}}\sin \theta e^{\pm i\varphi}, y_{2, \pm 2} = \sqrt{\dfrac{15}{32\pi}} \sin^2 \theta e^{\pm i2\varphi}$
$\Delta u = 0 \to u = \sum \limits_{l=0}^{\infty} \sum \limits_{m=-l}^{l}(c_lr^l + d_rr^{-(l+1)})y_{l,m}(\theta, \varphi)$

$(2)广义傅氏展开:$
$f(\theta, \varphi) = \sum \limits_{l=0}^{\infty} \sum \limits_{m=-l}^{l} C_{l,m}y_{l,m}(\theta, \varphi)$
$C_{l, m} = \int_0^{2\pi} \int_0^{\pi} f(\theta, \varphi) \overline{y_{l,m}(\theta, \varphi)} \sin \theta d \theta d \varphi$

$3例题：用球函数重新表示上述例题的解$
$u(r, \theta, \varphi) = \sum \limits_{l=0}^{\infty} \sum \limits_{m=-l}^{l} C_{l,m} r^l y_{l,m}(\theta, \varphi)$
$u|_{r=a} = u_0 \sin^2 \theta \cos 2 \varphi = \dfrac{u_0}{2} \sin^2 \theta [e^{i2 \varphi} + e^{-i2 \varphi}]$
$=\dfrac{u_0}{2}\sqrt{\dfrac{32\pi}{15}}[\sqrt{\dfrac{15}{32\pi}} \sin^2 \theta e^{i2 \varphi} + \sqrt{\dfrac{15}{32\pi}} \sin^2 \theta e^{-i2\varphi}]$
$= \sqrt{\dfrac{8\pi}{15}}u_0[y_{2, 2}(\theta, \varphi) +y_{2, -2}(\theta, \varphi)]$
$y_{2, \pm 2} = \sqrt{\dfrac{15}{32\pi}} \sin^2 \theta e^{\pm i 2 \varphi}$
$\therefore \sum \limits_{l=0}^{\infty} \sum \limits_{m=-l}^{l} C_{l,m} a^l y_{l,m}(\theta, \varphi) = \sqrt{\dfrac{8\pi}{15}}u_0[y_{2, 2}(\theta, \varphi) +y_{2, -2}(\theta, \varphi)]$
$\therefore C_{2,2}a^2 = C_{2,-2}a^2 = \sqrt{\dfrac{8\pi}{15}}u_0 即C_{2,\pm 2} = \sqrt{\dfrac{8 \pi}{15}} \dfrac{u_0}{a^2}, C_{l,m} \equiv 0(l \neq 2, m \neq \pm 2)$
$\therefore u(r, \theta, \varphi) = \sqrt{\dfrac{8\pi}{15}}u_0 \dfrac{r^2}{a^2}[y_{2,2}(\theta, \varphi) + y_{2, -2}(\theta, \varphi)]$

$\color{blue}{14.3.4 小结}$

$1.P_l(x):$
$(1)母函数关系式:$
$\dfrac{1}{\sqrt{1-2xt+t^2}} = \sum \limits_{l=0}^{\infty} P_l(x)t^l, |t| < 1 \quad (1)$
$(2)递推公式$
$A.(l+1)P_{l+1}(x) - (2l+1)xP_l(x) + lP_{l-1}(x) = 0 \quad (2)$
$B.(2l+1)P_l(x) = P_{l+1}^{\prime}(x) - P_{l-1}^{\prime}(x) \quad (3)$
$(3).正交性$
$\int_{-1}^{1}P_l(x)P_k(x)dx = \dfrac{2}{2l+1}\delta_{kl}, k,l = 0,1, 2\cdots (6)$
$(4)广义傅氏展开$
$f(x) = \sum \limits_{l=0}^{\infty} C_lP_l(x) \quad (9)$

$2.P_l^m(x):$
$(1) \left \lbrace \begin{array}{l}(1-x^2)y^{\prime \prime} - 2xy^{\prime} + [l(l+1) - \dfrac{m^2}{1-x^2}]y = 0 \\ y|_{x = \pm 1} \to 有限, \end{array} \right.$

$本征值:l(l+1), l = 0, 1, 2, \cdots$
$本征函数:y(x) = P_l^m(x) = (1-x^2)^{\frac{m}{2}} \dfrac{d^m}{dx^m} P_l(x), m = 0, 1, \cdots l \quad (6)$

$(2)(l+1-m)P_{l+1}^m(x) - (2l+1)xP_l^m(x) +(l+m)P_{l-1}^m(x) = 0 \quad (9)$
$(3)\int_{-1}^{1}P_l^m(x)P_k^m(x)dx = \dfrac{(l+m)!}{(l-m)!}\dfrac{2}{2l+1} \delta_{kl}$
$(4)f(x) =\sum \limits_{l=0}^{\infty}C_l^mP_l^m(x), C_l^m = \dfrac{(l-m)!}{(l+m)!}\dfrac{2l+1}{2}\int_{-1}^{1}f(x) P_l^m(x)dx$

$3.y_{l,m}(\theta, \varphi):$
$y_{l,m}(\theta, \varphi) = \sqrt{\dfrac{2l+1}{4\pi}\dfrac{(l-|m|)!}{(l+|m|)!}} P_l^{|m|} (\cos \theta) e^{im\varphi}$
$\Delta u = 0 \to u = \sum \limits_{l=0}^{\infty} \sum \limits_{m=-l}^{l} (c_lr^l + d_l r^{-(l+1)}) y_{l,m}(\theta, \varphi)$
$\int_0^{\pi} \int_0^{2\pi} y_{l,m}(\theta, \varphi) \overline{y_{l,n}(\theta, \varphi)}\sin \theta d \varphi d \theta = \delta_{kl} \delta_{nm}$
$f(\theta, \varphi) = \sum \limits_{l=0}^{\infty} \sum \limits_{m=-l}^{l} C_{l,m}y_{l,m}(\theta, \varphi)$
$C_{l,m} = \int_0^{2\pi} \int_0^{\pi} f(\theta, \varphi) \overline{y_{l,m}(\theta, \varphi)} \sin \theta d \theta d \varphi$