(一)投影算符和单位算符
1.取任意力学量算符的分立本征矢集合:
\[\mathinner{| i \rangle}\quad i\in Z^+
\]
??有投影算符pi:
\[\hat{p}_i=\mathinner{| i \rangle}\mathinner{\langle i |}
\\\space\\Rarr\forall \mathinner{| \psi \rangle}=\sum_j c_j \mathinner{| j \rangle}
\\\space\\begin{aligned}
\Rarr\hat{p}_i\mathinner{| \psi \rangle}&=\sum_jc_j\mathinner{|i \rangle}\mathinner{\langle i|j \rangle}\&=\sum_jc_j\mathinner{| i \rangle}\delta_{ij}\&=c_i\mathinner{| i \rangle}\quad\text{投影到}\mathinner{| i \rangle}\text{方向}
\end{aligned}
\\\space\take\quad\hat{I}=\sum_i\mathinner{| i \rangle}\mathinner{\langle i |}
\\\space\\begin{aligned}
\Rarr\hat{I}\mathinner{| \psi \rangle}&=\sum_i{\mathinner{| i \rangle}(\sum_j{c_j\mathinner{\langle i|j \rangle}})}\&=\sum_i c_i \mathinner{| i \rangle}=\mathinner{| \psi \rangle}\quad{单位算符}
\end{aligned}
\]
??对于任意一组基矢,满足单位算符等式的称为基矢的封闭性或完备性。例如,对于二维直角坐标系:
\[\mathinner{| 1 \rangle}=
\begin{pmatrix}
1\0
\end{pmatrix}\quad
\mathinner{| 2 \rangle}=
\begin{pmatrix}
0\1
\end{pmatrix}
\\\space\\Rarr \sum_i\mathinner{| i \rangle}\mathinner{\langle i |}=
\begin{pmatrix}
1\0
\end{pmatrix}
\begin{pmatrix}
1&0
\end{pmatrix}+
\begin{pmatrix}
0\1
\end{pmatrix}
\begin{pmatrix}
0&1
\end{pmatrix}
=\begin{pmatrix}
1 & 0\0 & 1
\end{pmatrix}
\]
2.考虑函数情形,先考虑x是离散的n个值,函数值为fn(xi):
\[\mathinner{| f_n \rangle}=
\begin{pmatrix}
f_n(x_1)\f_n(x_2)\...\f_n(x_n)
\end{pmatrix}\quad
\mathinner{| x_i \rangle}=\begin{pmatrix}
0\...\0\1\0\...\0
\end{pmatrix}\rarr\text{第i行}
\]
??易得如下结论:
\[\mathinner{| x_i \rangle}\mathinner{\langle x_i |}=\hat{p}_i\quad\sum_i\mathinner{| x_i \rangle}\mathinner{\langle x_i |}=\hat{I}\quad\mathinner{\langle x_i|x_j \rangle}=\delta_{ij}
\\\space\\begin{aligned}
&\mathinner{| f_n \rangle}
&&=
&&\sum_i\mathinner{| i \rangle}\mathinner{\langle i | f_n \rangle}\&f_n(x_i)\mathinner{| x_i \rangle}
&&=
&&\mathinner{| x_i \rangle}\mathinner{\langle x_i | f_n \rangle}\&\mathinner{\langle f_n |f_n \rangle}
&&=
&&\mathinner{\langle f_n | \hat{I} | f_n \rangle} \&&&=
&& \sum_i|f(x_i)|^2\&\mathinner{\langle f_n |g_n \rangle}
&&=
&&\mathinner{\langle f_n | \hat{I} | g_n \rangle} \&&&=
&& \sum_i f^*(x_i)g(x_i)\&\mathinner{| f_n \rangle}+\mathinner{|g_n \rangle}
&&=
&&\hat{I}(\mathinner{| f_n \rangle}+\mathinner{|g_n \rangle})\&&&=
&& \sum_i[f(x_i)+g(x_i)]\mathinner{| i \rangle}\&&&=
&& \begin{pmatrix}
f_n(x_1)+g_n(x_1)\f_n(x_2)+g_n(x_2)\...\f_n(x_n)+g_n(x_n)
\end{pmatrix}\\end{aligned}
\]
??那么,容易拓展到x取R的情况:
\[\begin{aligned}
f(x)&=\int\delta(x-x‘)f(x‘)dx‘\&=\int\mathinner{\langle x|x‘ \rangle}\mathinner{\langle x‘|f \rangle}dx‘\&=\mathinner{\langle x|\hat{I}|f\rangle}=\mathinner{\langle x|f \rangle}
\end{aligned}
\\\space\as \quad \hat{I}=\int\mathinner{|x‘ \rangle}\mathinner{\langle x‘|}dx‘=\int\mathinner{|x \rangle}\mathinner{\langle x|}dx
\\\space\\begin{aligned}
\Rarr\mathinner{\langle f|g \rangle}
&=\mathinner{\langle f| \hat{I} |g \rangle}\&=\int\mathinner{\langle f|x \rangle}\mathinner{\langle x|g \rangle}dx\&=\int{f^*(x)g(x)}dx
\end{aligned}
\\\space\take\quad\hat{p}_{x_0}=\mathinner{|x_0 \rangle}\mathinner{\langle x_0|}
\\\space\\begin{aligned}
\Rarr\mathinner{\hat{p}_{x_0}|f \rangle}
&=\mathinner{| x_0 \rangle \langle x_0 |f \rangle}\&=f(x_0)\mathinner{| x_0 \rangle}
\end{aligned}
\]
(二)格兰-施密特正交化与勒让德多项式
??设一组向量线性独立而不正交:
\[\mathinner{| i \rangle}\quad i\in Z^+
:\sum_i c_i{\mathinner{| i \rangle}}=\not 0\quad\exists{i_1=\not i_2}\Rarr\mathinner{\langle i_1|i_2 \rangle}=\not0
\]
??那么有格兰施密特正交化方法:
\[\begin{aligned}
\mathinner{| v_i \rangle}
&=\mathinner{|i \rangle}-\sum_{n=1}^{i-1}\frac{\mathinner{\langle i|n \rangle}}{\mathinner{\langle n|n \rangle}}\mathinner{| n \rangle}\\mathinner{| u_i \rangle}&=\frac{\mathinner{|v_i \rangle}}{\sqrt{\mathinner{\langle v_i|v_i \rangle}}}
\end{aligned}
\]
??这等效于如下的过程(减去平行分量):
\[\begin{aligned}
\mathinner{|v_1 \rangle}&=\frac{\mathinner{|1 \rangle}}{\mathinner{\langle 1|1 \rangle}}\\mathinner{| v_i \rangle}&=\mathinner{|i \rangle}-\sum_{n=1}^{i-1}\hat{p}_{v_n}\mathinner{|i \rangle}\\&=\mathinner{|i \rangle}-\sum_{n=1}^{i-1}\mathinner{|v_n \rangle}\mathinner{\langle v_n|i \rangle}\\mathinner{| u_i \rangle}&=\frac{\mathinner{|v_i \rangle}}{\sqrt{\mathinner{\langle v_i|v_i \rangle}}}
\end{aligned}
\]
??取泰勒展开式中x幂次为基矢:
\[\mathinner{| i \rangle}=x^{i-1}\quad i\in Z^+
\]
??以[-1,1]为积分区间,对各基矢正交归一化,有:
\[\begin{aligned}
\mathinner{| v_1 \rangle}&=\frac{1}{\sqrt{2}}\\mathinner{| v_2 \rangle}&=x\\mathinner{| v_3 \rangle}&=x^2-\frac{1}{3}\&...\\Rarr\mathinner{| u_1 \rangle}&=\frac{1}{\sqrt2}\\mathinner{| u_2 \rangle}&=\sqrt{\frac{2}{3}}x\\mathinner{| u_3 \rangle}&=\sqrt{\frac{5}{8}}(x^2-\frac{1}{3})\&...
\end{aligned}\\]
??un(x)即为勒让德多项式的集合。
(三)求特征向量的另两种方法
1.陶哲轩方法(适用于计算机)
??设n阶矩阵A有特征值λi,其对应特征向量vi的第j个分量有如下关系式:
\[|v_{ij}|^2\prod_{k=1;k=\not i}^n (\lambda_i-\lambda_k)=\prod_{k=1}^{n-1}(\lambda_i-\mu_{k})
\]
??其中μkj是A去掉第j行第j列元素后的主子矩阵Mj(易证这个矩阵必定厄米)的第k个特征值。设二阶矩阵:
\[A=\begin{pmatrix}
\mu_2 & x\x^* & \mu_1
\end{pmatrix}\Rarr Av_i=\lambda_i v_i \quad i=1,2\\\space\\Rarr M_1=\begin{pmatrix} \mu_1 \end{pmatrix}\quad M_2=\begin{pmatrix} \mu_2 \end{pmatrix}\quad\text{特征值显然}\\\space\\begin{aligned}
\Rarr |v_{11}|^2&=\frac{\lambda_1-\mu_1}{\lambda_1-\lambda_2}\|v_{12}|^2&=\frac{\lambda_1-\mu_2}{\lambda_1-\lambda_2}\|v_{21}|^2&=\frac{\lambda_1-\mu_1}{\lambda_2-\lambda_1}\|v_{22}|^2&=\frac{\lambda_1-\mu_2}{\lambda_2-\lambda_1}\\end{aligned}
\]
2.旋转基矢法(二维实矩阵)
??取二维实矩阵Ω和一个旋转矩阵Ut:
\[\Omega=
\begin{pmatrix}
\Omega_{11}&\Omega_{12}\\Omega_{21}&\Omega_{22}
\end{pmatrix}
\quad\Omega_{12}=\Omega_{21}
\\\space\U^t=\begin{pmatrix}
\cos\theta&\sin\theta\-\sin\theta&\cos\theta
\end{pmatrix}
\\\space\\Rarr U^t\Omega U=\begin{pmatrix}
\cos^2\theta\Omega_{11}+\sin^2\theta\Omega{22}+\Omega_{12}\sin{2\theta}& 2\Omega_{12}\sin{2\theta}-(\omega_{11}-\Omega_{22})\cos{2\theta}\2\Omega_{12}\sin{2\theta}-(\omega_{11}-\Omega_{22})\cos{2\theta}&
\sin^2\theta\Omega_{11}+\cos^2\theta\Omega{22}+\Omega_{12}\sin{2\theta}
\end{pmatrix}
\\\space\\overset{对角元为0}{\Rarr}\theta_0=\frac{1}{2}\arctan\frac{2\Omega_{12}}{\Omega_{11}-\Omega_{22}}
\\\space\\Rarr
\mathinner{| v_1 \rangle}=\begin{pmatrix}
\cos\theta_0\\sin\theta_0
\end{pmatrix}
\quad
\mathinner{| v_2 \rangle}=\begin{pmatrix}
-\sin\theta_0\\cos\theta_0
\end{pmatrix}
\]
(四)平面波与高斯函数
??对于一个体系的波函数(未知,通常不能通过求解薛定谔方程得到),可以按某种基矢展开,然后利用变分法的思想,通过线性优化得到收敛的近似波函数。
??在从头算理论中,基矢一般是高斯波函数;而在第一性原理中,基矢一般是平面波函数。
??关于平面波函数的傅里叶展开的狄拉克符号写法:
\[k(x)=\mathinner{\langle x|k \rangle}\overset{def}{=}\frac{1}{\sqrt{2\pi}}e^{ikx}\\\space\x(k)=\mathinner{\langle k|x \rangle}\overset{def}{=}\frac{1}{\sqrt{2\pi}}e^{-ikx}\\\space\\Rarr\int\mathinner{|k \rangle}\mathinner{\langle k|}dk=\hat{I}_k
\\\space\\begin{aligned}
\Rarr f(k)&=\mathinner{\langle k|f \rangle}\&=\mathinner{\langle k | \hat{I}_x | f \rangle}\&=\int_\infty\mathinner{\langle k | x \rangle}\mathinner{\langle x | f \rangle}dx\&=\frac{1}{\sqrt{2\pi}}\int_\infty f(x)e^{-ikx}dx\f(x)&=\mathinner{\langle x|f \rangle}\&=\mathinner{\langle k | \hat{I}_k | f \rangle}\&=\int_\infty\mathinner{\langle x | k \rangle}\mathinner{\langle k | f \rangle}dx\&=\frac{1}{\sqrt{2\pi}}\int_\infty f(k)e^{ikx}dx\\end{aligned}\\\space\\begin{aligned}
\Rarr f(x)&=\frac{1}{2\pi}\int_\infty \int_\infty f(x‘)e^{-ikx‘}dx‘e^{ikx}dk\&=\int_\infty f(x‘)\frac{1}{2\pi}\int_\infty e^{ik(x-x‘)}dkdx‘\&=\int_\infty f(x‘)\delta(x-x‘)dx‘
\end{aligned}
\\\space\\overset{同理}{\Rarr}\frac{1}{2\pi}\int_\infty
e^{i(k-k‘)x}dx=\delta(k-k‘)
\]
??