初学量子,对一个问题的思考,请教老师指正

新用户注册 | 用户登陆 | 刷新
论坛嘉宾: sage

Quit


发表文章数: 22
内力值: 89/89
贡献度: 190
人气: 164

初学量子,对一个问题的思考,请教老师指正 [文章类型: 原创]

初学量子力学,有很多问题不清楚,想向大家请教,请大家不吝指正,3Q first

2个月之前,刚开始学量子的线性代数基础的时候,看到一个题,当时百思不得其解,郁闷了颇久
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
假设A和B都是自共轭算子,并且二者的对易式为:[A,B]=i.
设:A|ψ〉=a|ψ〉,即|ψ〉是A的本征值为a 的本征态;并且|ψ〉已经归一化:〈ψ| ψ〉=1.
我们现在来求〈ψ|[A,B]ψ〉=〈ψ|ABψ〉-〈ψ|BAψ〉.

因为[A,B]=i,所以上式左边=〈ψ|[A,B]ψ〉=i〈ψ|ψ〉=i;
因为A是自共轭算子,所以〈ψ|ABψ〉=〈Aψ|Bψ〉=a〈ψ|Bψ〉,而〈ψ|BAψ〉=a〈ψ|Bψ〉,因此上式右边=0.
难道矛盾?
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

现在过了一段时间来看,好象有点发现问题出在什么地方了,但是不知道考虑得对不对,请教一下大家
(因为数学水平确实有限,想的东西里面欠缺了很多严格的数学证明,所以不敢说是对这个问题的试图解释,只是想到的一些可能性)

:

如果:题目条件确实是〈ψ| ψ〉归一化到〈ψ| ψ〉=i
那么这个态空间就是严格的Hilbert空间,因为| ψ〉是A在a的本征矢 这个条件对| ψ〉的限制,满足如上条件的| ψ〉就只能是零矢| 0〉,但是这样的话A的本征矢就只能是|0>,这与题目中A为厄米这个前提矛盾,于是在严格的Hilbert空间里面不存在两个厄米算符对易=i
(但是还有一种可能性,即这个Hilbert空间是0维的0空间,只有零矢量|0>存在于这个空间,这个时候A倒真正成了厄米算符,在这个极端的情况下,满足题目的要求)


但是如果我们的态空间是物理的Hilbert空间,那么它还应该包含一些广义函数在内
在这个态空间里面,A如果是厄米算符,那么它的本征矢就应该归一化到〈ψ'| ψ〉=δ(a'-a) (|ψ'>和|ψ>分别为A在a'和a的本征矢)
在这个前提下,再对上面这个问题进行分析,就是:
左边= 〈ψ|[A,B]ψ〉 = 〈ψ|i|ψ〉=i<ψ|ψ>=i〈ψ'| ψ〉|a'=a=iδ(a'-a) |a'=a
右边= 〈ψ|(AB-BA)|ψ〉=〈ψ|AB|ψ〉-<ψ|BA|ψ>
= 〈ψ'|AB|ψ〉-<ψ'|BA|ψ> |a'=a
= a'〈ψ'|B|ψ〉- a<ψ'|B|ψ> |a'=a
= (a'-a)〈ψ'|B|ψ〉|a'=a
将a'考虑成一个变量,则 〈ψ'|B|ψ〉同"〈ψ'| ψ〉=δ(a'-a)" 类似,也是一个广义函数,把它定义为〈ψ'|B|ψ〉≡B(a',a),这个B(a',a)就是一个关于a'的广义函数

那么以上的左和右都是关于广义函数的运算的问题了,不能再按照常规方式来处理,以下试图在广义函数的定义下处理这个问题:
因为左边=iδ(a'-a) |a'=a,右边(a'-a)〈ψ'|B|ψ〉|a'=a=(a'-a)B(a'-a) |a'=a,他们看上去都是广义函数在a'=a点的"函数值"的等式,但是因为广义函数是在线性泛函意义下定义的,我们不能讨论他们的函数值

于是我们就只能在泛函意义下讨论iδ(a'-a)和(a'-a)B(a'-a) 这两个函数相等的问题
由于我的数学水平确实有限,用恒等变形把B(a',a)的表达式得到,再证明在广义函数的意义下左右可以恒等,这个任务对我来说确实太困难了,但是因为[A,B]=i 提醒我想到了[X,P]=i ,而在坐标算符的表象下P的形式为-i(d/dx) ,即∫P(x',x)ψ(x)dx=pψ(x'),∫P(x',x)ψ(x)dx=-i(d/dx')ψ(x'),由δ函数的导数的定义(δ'(x),f)=-(δ(x),f'),可以推导出<x'|P|x>=P(x',x)=i(d/dx)δ(x'-x)
这里X算符和P算符的关系与题目中A算符和B算符的关系完全是一样的,而且我们要的那个B(a',a)=〈ψ'|B|ψ〉也与P(x',x)=<x'|P|x>一样,是满足对易=i的两个算符中的一个算符插入到另外一个算符表象的归一化关系中得到的函数。
所以把X,P带回到原来的的题目中间去,也完全符合题设条件,只是用的符号不同而已,于是我们可以完全写成B(a',a)=-i(d/da')δ(a'-a)
现在再来比较原式的左右:
左边=iδ(a'-a)
右边=(a'-a)i(d/da)δ(a'-a)

在广义函数的意义下,有(δ'(x),f)=-(δ(x),f'),所以 (右,f)=(-(a'-a)i(d/da')δ(a'-a),f)=-(δ(a'-a),-i[(a'-a)f]')=i[(a'-a)f]'|a'=a=i[f(a')+(a'-a)f'(a')] |a'=a=i[f(a)+0f'(a)]=if(a)
;(左,f)=(iδ(a'-a) ,f)=if(a')
所以在广函定义下,得到证明:左边=右边
我们原来的式子是左右的广义函数在a'=a点的函数值,于是也可以得到证明是相等的,所以矛盾得到消除,原来的两种求〈ψ|[A,B]ψ〉的方法得到的答案是相同的。

以上就证明了在物理的Hilbert空间中,至少存在一对算符X和P,满足如上的算符A和B的要求。但是在这个空间中,坐标算符X和动量算符P是否是唯一一对满足这个要求的算符对,这个问题并没有得到证明,但是这个"唯一性"的证明对消除题目中看似存在的矛盾,不是必须的。
(实际上,坐标算符X,和动量算符P在我们的态空间的存在性,在这里也只是在物理意义上直接拿下来的,严格的说,它们是需要严格的数学证明的)








其实我觉得这个问题的要结在于:
我们的题设条件
一,A和B为厄米算符
二,[A,B]=i
三,A|ψ>=|ψ>a

对算符A,B和矢|ψ>都是做了限定的
对于基归一到δij的Hilbert空间,这个限定表现为对|ψ>限定为|0>,既而对空间和算符都产生了限定
对于归一化到δ函数的物理Hilbert空间,这个限定表现为对算符A,B的限定


其实这道题目隐藏的更深的一个问题是,题目应该给出讨论的矢量空间是什么空间这个前提条件

<x|F>=<x|∫|p><p|F>dp=∫<x|p><p|F>dp
<p|F>=<p|∫|x><x|F>dx=∫<p|x><x|F>dx
原来这就是傅立叶变换,我又被忽悠了:(

发表时间: 2006-09-27, 11:58:26 个人资料

季候风


发表文章数: 262
内力值: 310/310
贡献度: 3398
人气: 154

论坛嘉宾学术成员

Re: 初学量子,对一个问题的思考,请教老师指正 [文章类型: 原创]

看来你知道广义函数但是还不知道无界算子

Stone - von Neumann 定理:满足 Heisenberg 对易关系的算子只能是 q 和 i d/dq (作用在所有 q 的平方可积函数组成的 Hilbert 空间)。

这两个算子没有离散本征值,也就是你说的,没有真正在 Hilbert 空间里的本征矢量。当然我们可以用广义函数,这也是在物理上最方便的办法,但是严格的数学理论要用到 von Neumann 关于无界算子谱分解的理论。

如果要考虑 q 的算子函数,那么在数学上还需要 Gelfand 关于 Banach 代数的理论。

发表时间: 2006-09-27, 18:04:12 个人资料

季候风


发表文章数: 262
内力值: 310/310
贡献度: 3398
人气: 154

论坛嘉宾学术成员

Re: 初学量子,对一个问题的思考,请教老师指正 [文章类型: 原创]

不好意思,在 Stone - von Neumann 定理里漏了 “厄米” 两个字 --- 数学上严格来说是 “自伴” 两个字

发表时间: 2006-09-27, 18:06:26 个人资料

laworder


发表文章数: 89
内力值: 139/139
贡献度: 1863
人气: 91

学术成员

Re: 初学量子,对一个问题的思考,请教老师指正 [文章类型: 原创]

Actually, the whole thing can be understood without so much math.

here is an elementary explanation:

(1) Notice that the commutator of any two operators HAS TO BE AN OPERATOR:
[A,B]=iC
(even though C can be an identity operator. Expressions like "[q,p]=i" are the root of confusion for the beginners.)

(2) In the eigen state, |φ_A>, of operator A, the average of [A,B] IS zero:

left= <φ_A|[A,B]|φ_A> = 0
right = <φ_A|C|φ_A> = <φ_A|C Σ_C C_ac |φ_C> = 0

(3) In any state |ψ>, the average of [A,B] is

lEFT = <ψ|[A,B]| ψ> =Σ_a’ C^*_a’ <φ_A’|[A,B] Σ_a C_a |φ_A>
=(Σ_a,a’ C^*_a’C_a a’-Σ_a,a’ C^*_a’C_a a) <φ_A’|B|φ_A>
=Σ_a,a’ C^*_a’C_a (a’-a) <φ_A’|B|φ_A>
which may be nonzero.

Or

RIGHT = <ψ|C| ψ> =Σ_a’ C^*_a’ <φ_A’|C Σ_a C_a |φ_A>
=Σ_a,a’ C^*_a’C_a <φ_A’|C|φ_A>
which, again, may be nonzero.

人能理解世界是最大的科学之谜。

发表时间: 2006-09-28, 01:14:49 个人资料

季候风


发表文章数: 262
内力值: 310/310
贡献度: 3398
人气: 154

论坛嘉宾学术成员

Re: 初学量子,对一个问题的思考,请教老师指正 [文章类型: 原创]

<φ_A|C Σ_C C_ac |φ_C> = 0
~~~~~~~~~~~~~~~~

why?

发表时间: 2006-09-28, 11:12:36 个人资料

laworder


发表文章数: 89
内力值: 139/139
贡献度: 1863
人气: 91

学术成员

Re: 初学量子,对一个问题的思考,请教老师指正 [文章类型: 原创]

<φ_A|C Σ_C C_ac |φ_C> = 0
~~~~~~~~~~~~~~~~

why?
~~~~~~~~~~~~~~~~~~~~~~~~


Are you kidding, Professor Ji Hou Feng?

Anyway, C_ac = <φ_A |φ_C> = 0 and C_bc = <φ_B |φ_C> = 0 if [A,B] = iC.

[A,B] = iC means C is orthogonal to A and B:

Tr(CA) = Tr ([A,B]A) = Tr (ABA-BAA) = 0.
Tr(CB) = Tr ([A,B]B) = Tr (ABB-BAB) = 0.

人能理解世界是最大的科学之谜。

发表时间: 2006-09-29, 10:49:22 个人资料

sage


发表文章数: 359
内力值: 334/334
贡献度: 5130
人气: 237

客栈长老论坛嘉宾学术成员

Re: 初学量子,对一个问题的思考,请教老师指正 [文章类型: 原创]

>Anyway, C_ac = <φ_A |φ_C> = 0 and C_bc = <φ_B |φ_C> = 0 if [A,B] = iC.

I guess I am confused about this statement as well. Are you saying if

[A,B]=iC for any operator A, B and C, the eigenstates of A, B and C are orthogonal to each other? This is not true, for example, consider the angular momentum operator.

发表时间: 2006-09-29, 11:18:23 个人资料

季候风


发表文章数: 262
内力值: 310/310
贡献度: 3398
人气: 154

论坛嘉宾学术成员

Re: 初学量子,对一个问题的思考,请教老师指正 [文章类型: 原创]

如果能对它们取 trace, [A,B]= i 这个式子本身就已经崩溃了.....

发表时间: 2006-09-29, 11:32:45 个人资料

laworder


发表文章数: 89
内力值: 139/139
贡献度: 1863
人气: 91

学术成员

Re: 初学量子,对一个问题的思考,请教老师指正 [文章类型: 混合]

To sage:

I guess I am confused about this statement as well. Are you saying if
[A,B]=iC for any operator A, B and C, the eigenstates of A, B and C are orthogonal to each other? This is not true, for example, consider the angular momentum operator.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

No problem. For angular momentum operators, I_i, I_j, i,j=x,y,z, i≠j, one always has Tr(I_iI_j) = 0, Or, equivalently, <m_i|m_j> = 0 if i≠j where |m_i> and |m_j> are the eigenstates of I_i and I_j, respectively.


To Ji Hou Feng:
如果能对它们取 trace, [A,B]= i 这个式子本身就已经崩溃了.....
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

As stated in my first message related to this question, [A,B]= iC, rather than [A,B]= i, should be used.

Tr(A^{+} B) = 0 is the definition of orthogonality of two operators.

人能理解世界是最大的科学之谜。

发表时间: 2006-09-30, 03:05:57 个人资料

laworder


发表文章数: 89
内力值: 139/139
贡献度: 1863
人气: 91

学术成员

Re: 初学量子,对一个问题的思考,请教老师指正 [文章类型: 混合]

Oops, sorry, an error occurred in haste:

Or, equivalently, <m_i|m_j> = 0 if i≠j where |m_i> and |m_j> are the eigenstates of I_i and I_j, respectively.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

should be

Or, equivalently, <m_i|I_j|m_i> = 0 if i≠j where |m_i> and |m_j> are the eigenstates of I_i and I_j, respectively.

人能理解世界是最大的科学之谜。

发表时间: 2006-09-30, 03:23:54 个人资料

季候风


发表文章数: 262
内力值: 310/310
贡献度: 3398
人气: 154

论坛嘉宾学术成员

Re: 初学量子,对一个问题的思考,请教老师指正 [文章类型: 原创]

As stated in my first message related to this question, [A,B]= iC, rather than [A,B]= i, should be used.

Tr(A^{+} B) = 0 is the definition of orthogonality of two operators.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

我们现在谈论的是 Heisenberg 对易关系, 你的 C 就是 identity, [p,q]=i (id)
如果它们能用矩阵表示而且能取 trace, 那么根据对易关系本身, tr ([p,q]) = i tr(id),
但这是一个矛盾, 因为左边为0, 右边不为零.

既然不能取 trace, 你拿什么定义算子的 "正交" ? 再说算子的 "正交" 一般只对投影算子而言.

发表时间: 2006-09-30, 11:43:45 个人资料

Quit


发表文章数: 22
内力值: 89/89
贡献度: 190
人气: 164

Re: 初学量子,对一个问题的思考,请教老师指正 [文章类型: 原创]

<φ_A|C Σ_C C_ac |φ_C> = 0

——————————————————————————————————————
符号实在没有看懂*_!
但是[a,b]=i确实不是我原先confused的地方,[a,b]=i我是当成数乘算符来看的。

<x|F>=<x|∫|p><p|F>dp=∫<x|p><p|F>dp
<p|F>=<p|∫|x><x|F>dx=∫<p|x><x|F>dx
原来这就是傅立叶变换,我又被忽悠了:(

发表时间: 2006-09-30, 17:08:51 个人资料

Quit


发表文章数: 22
内力值: 89/89
贡献度: 190
人气: 164

Re: 初学量子,对一个问题的思考,请教老师指正 [文章类型: 原创]

看来你知道广义函数但是还不知道无界算子

Stone - von Neumann 定理:满足 Heisenberg 对易关系的算子只能是 q 和 i d/dq (作用在所有 q 的平方可积函数组成的 Hilbert 空间)。

这两个算子没有离散本征值,也就是你说的,没有真正在 Hilbert 空间里的本征矢量。当然我们可以用广义函数,这也是在物理上最方便的办法,但是严格的数学理论要用到 von Neumann 关于无界算子谱分解的理论。

如果要考虑 q 的算子函数,那么在数学上还需要 Gelfand 关于 Banach 代数的理论。
——————————————————————————————————————————————

小可数学基础确实顶不住,现在学量子和它所用到的数学基础全部都是零散地自学的,更多的东西就是完全没有听说过了:(

<x|F>=<x|∫|p><p|F>dp=∫<x|p><p|F>dp
<p|F>=<p|∫|x><x|F>dx=∫<p|x><x|F>dx
原来这就是傅立叶变换,我又被忽悠了:(

发表时间: 2006-09-30, 17:13:47 个人资料

季候风


发表文章数: 262
内力值: 310/310
贡献度: 3398
人气: 154

论坛嘉宾学术成员

Re: 初学量子,对一个问题的思考,请教老师指正 [文章类型: 原创]

我建议你不要纠缠在这些数学基础上面, 而是要充分理解符号后面的物理意义

发表时间: 2006-09-30, 21:14:43 个人资料

西门吹牛


发表文章数: 312
内力值: 301/301
贡献度: 1790
人气: 472

客栈长老学术成员

Re: 初学量子,对一个问题的思考,请教老师指正 [文章类型: 原创]

难道我学的东西跟laworder学的东西不是同一个东西?怎么我看不懂啊

一舞剑气动四方,天下英雄莫能挡
形踪飘忽疑无影,冷面郎君傲雪霜

发表时间: 2006-10-01, 03:33:06 个人资料

laworder


发表文章数: 89
内力值: 139/139
贡献度: 1863
人气: 91

学术成员

Re: 初学量子,对一个问题的思考,请教老师指正 [文章类型: 混合]

Expressions like “[q,p]=i”, “[x,p_x]=i h_bar” are confusing, to say the least, and WRONG to say the most. (Unfortunately there are so many similar slopy equations in the textbooks of quantum mechanics)

The correct expression should be written as (I don’t claim originality :-):

[q_m, p^n] = iδ_m ^n or [q_m, p^n] = ih_bar δ_m ^n

because:

(1) p is defined in the dual space of q, and vice versa. In physics terms, q is covariant and p is countervariant.

(2) It captures the essence of uncertainty relation: the simultaneous precise measurement of an observable and its dual (component) is impossible. It is simply what Fourier did ~100 years before quantum mechanics.

In certain sense, therefore, uncertainty has nothing to do with quantum mechanics. It is a universal (purely mathematical) fact. “Quantum” comes in when one merely applies de Brodgile’s genius idea p=hk to Fourier’s (uncertainty) relation. That’s why I always call it “Fourier-de Brodgile relation” (I do claim originality :-) or more generously “Fourier-de Brodgile-Heisenberg relation” although I don’t see anything substantial Heisenberg did. Perhaps that’s why Einstein never mentioned that guy.

With above correct expression of uncertainty relation, one can safely use the good old elementary matrix algebra without worrying about “the peculiar properties of infinite dimensional spaces”, “the boundless spectra of p and q”, “be aware the spectra here are continuous”, “don’t take trace here” bla bla. The beginners not only get the correct physics picture at once but also are not scared by terrorist remarks like “you need to learn more functional analysis, or Heisenberg algebra, to really understand this subtle point”. (I don't mean functional analysis or geometric algebra etc are useless).

人能理解世界是最大的科学之谜。

发表时间: 2006-10-01, 17:27:28 个人资料

季候风


发表文章数: 262
内力值: 310/310
贡献度: 3398
人气: 154

论坛嘉宾学术成员

Re: 初学量子,对一个问题的思考,请教老师指正 [文章类型: 原创]

Expressions like “[q,p]=i”, “[x,p_x]=i h_bar” are confusing, to say the least, and WRONG to say the most. (Unfortunately there are so many similar slopy equations in the textbooks of quantum mechanics)

The correct expression should be written as (I don’t claim originality :-):

[q_m, p^n] = iδ_m ^n or [q_m, p^n] = ih_bar δ_m ^n
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

这跟 [q,p]=i 的唯一区别就是经典相空间的维数 --- 而且这里的指标跟 p, q 作为矩阵的指标完全是两回事..........


because:

(1) p is defined in the dual space of q, and vice versa. In physics terms, q is covariant and p is countervariant.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
恰恰相反, p 是协变的, 而 q 只是坐标, 根本谈不上协变反变



(2) It captures the essence of uncertainty relation: the simultaneous precise measurement of an observable and its dual (component) is impossible. It is simply what Fourier did ~100 years before quantum mechanics.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~
看来你才真有泛数学化的倾向嘛, hoho


In certain sense, therefore, uncertainty has nothing to do with quantum mechanics. It is a universal (purely mathematical) fact. “Quantum” comes in when one merely applies de Brodgile’s genius idea p=hk to Fourier’s (uncertainty) relation. That’s why I always call it “Fourier-de Brodgile relation” (I do claim originality :-) or more generously “Fourier-de Brodgile-Heisenberg relation” although I don’t see anything substantial Heisenberg did. Perhaps that’s why Einstein never mentioned that guy.

With above correct expression of uncertainty relation, one can safely use the good old elementary matrix algebra without worrying about “the peculiar properties of infinite dimensional spaces”, “the boundless spectra of p and q”, “be aware the spectra here are continuous”, “don’t take trace here” bla bla.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
逻辑混乱 --- Fourier 分析跟 "old elementary matrix algebra" 有什么关系? 就着你自己的话说, 就是因为有 Fourier 分析的数学背景, 我们才需要知道一些泛函分析.


The beginners not only get the correct physics picture at once but also are not scared by terrorist remarks like “you need to learn more functional analysis, or Heisenberg algebra, to really understand this subtle point”. (I don't mean functional analysis or geometric algebra etc are useless).
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
学点已经成熟了快一个世纪的理论是什么了不得的事情?

发表时间: 2006-10-01, 17:53:24 个人资料

laworder


发表文章数: 89
内力值: 139/139
贡献度: 1863
人气: 91

学术成员

Re: 初学量子,对一个问题的思考,请教老师指正 [文章类型: 混合]

Expressions like “[q,p]=i”, “[x,p_x]=i h_bar” are confusing, to say the least, and WRONG to say the most. (Unfortunately there are so many similar slopy equations in the textbooks of quantum mechanics)

The correct expression should be written as (I don’t claim originality :-):

[q_m, p^n] = iδ_m ^n or [q_m, p^n] = ih_bar δ_m ^n
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

这跟 [q,p]=i 的唯一区别就是经典相空间的维数 --- 而且这里的指标跟 p, q 作为矩阵的指标完全是两回事..........

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Instances of the matrix expressions of q and p: q_x _{x_1 x_2}, p^y ^{p_z1 p_z2}.


because:

(1) p is defined in the dual space of q, and vice versa. In physics terms, q is covariant and p is countervariant.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
恰恰相反, p 是协变的, 而 q 只是坐标, 根本谈不上协变反变
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
I acknowledge your point. It suffices by saying “p is defined in the dual space of q, and vice versa.”


(2) It captures the essence of uncertainty relation: the simultaneous precise measurement of an observable and its dual (component) is impossible. It is simply what Fourier did ~100 years before quantum mechanics.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~
看来你才真有泛数学化的倾向嘛, hoho
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Uncertainty IS mathematical. (I never accused anybody of “matholic”.)


In certain sense, therefore, uncertainty has nothing to do with quantum mechanics. It is a universal (purely mathematical) fact. “Quantum” comes in when one merely applies de Brodgile’s genius idea p=hk to Fourier’s (uncertainty) relation. That’s why I always call it “Fourier-de Brodgile relation” (I do claim originality :-) or more generously “Fourier-de Brodgile-Heisenberg relation” although I don’t see anything substantial Heisenberg did. Perhaps that’s why Einstein never mentioned that guy.

With above correct expression of uncertainty relation, one can safely use the good old elementary matrix algebra without worrying about “the peculiar properties of infinite dimensional spaces”, “the boundless spectra of p and q”, “be aware the spectra here are continuous”, “don’t take trace here” bla bla.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
逻辑混乱 --- Fourier 分析跟 "old elementary matrix algebra" 有什么关系? 就着你自己的话说, 就是因为有 Fourier 分析的数学背景, 我们才需要知道一些泛函分析.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Don’t’ forget we have been discussing a problem posed by a “QM beginner” (I hope s/he didn’t mean to set someone up) . I just meant to pass a message to a QM beginner "don't worry, you can understand the essence of QM with matrix algebra". If I happened to hurt the feeling of some perfectionist expert, then I apologize.


The beginners not only get the correct physics picture at once but also are not scared by terrorist remarks like “you need to learn more functional analysis, or Heisenberg algebra, to really understand this subtle point”. (I don't mean functional analysis or geometric algebra etc are useless).
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
学点已经成熟了快一个世纪的理论是什么了不得的事情?

No comment.

人能理解世界是最大的科学之谜。

发表时间: 2006-10-01, 18:41:11 个人资料

sage


发表文章数: 359
内力值: 334/334
贡献度: 5130
人气: 237

客栈长老论坛嘉宾学术成员

Re: 初学量子,对一个问题的思考,请教老师指正 [文章类型: 原创]

>In certain sense, therefore, uncertainty has nothing to do with quantum mechanics. >It is a universal (purely mathematical) fact. “Quantum” comes in when one merely >applies de Brodgile’s genius idea p=hk to Fourier’s (uncertainty) relation.

===============================================================================

No. The point of uncertainty principle is its interpretation, which rooted in the theory of measurement of quantum mechanics. It is not mathematics. It is a deep statement of dynamics.

If it is only mathematics, there should be an uncertainty principle for classical mechanics as well.

发表时间: 2006-10-02, 00:02:28 个人资料

sage


发表文章数: 359
内力值: 334/334
贡献度: 5130
人气: 237

客栈长老论坛嘉宾学术成员

Re: 初学量子,对一个问题的思考,请教老师指正 [文章类型: 原创]

>Don’t’ forget we have been discussing a problem posed by a “QM beginner” (I >hope s/he didn’t mean to set someone up) . I just meant to pass a message to a QM >beginner "don't worry, you can understand the essence of QM with matrix algebra". >If I happened to hurt the feeling of some perfectionist expert, then I apologize.
-------------------------------------------------------------------------------

The message is probably correct. On the other hand, your explanation is probably wrong. Therefore, it weakens the message.

发表时间: 2006-10-02, 00:05:34 个人资料
您尚未登陆 | 用户登陆