0
$\begingroup$

I cannot understand the derivation in Louis Garbe article (https://arxiv.org/abs/1910.00604) about how to obtain the covariance matrix equation from Fokker-Planck equation for the Wigner function in Appendix_Dissipative dynamics.

The Lindblad equation above into a Fokker-Planck equation for the Wigner function: \begin{equation} \frac{\partial W}{\partial t}(x,p)=-\omega_0p\frac{\partial W}{\partial x} - \omega_0(Xg^2-1)x\frac{\partial W}{\partial p} + \kappa(2W + \sum_{i=1}^2x_i\partial_iW + \sum_{i,j=1}^2\partial_i\sigma^L_{ij}\partial_jW), \end{equation} where $x_1=x$, $x_2=p$,and $$\sigma^L =\frac{1}{2}\begin{bmatrix}1 & 0 \\ 0 & 1+Xg^2\frac{\Gamma\omega_0}{\Omega\kappa}\end{bmatrix}$$

In article, an author say "Since this equation is quadratic in x and p, it can be solved by a Gaussian ansatz $ W=\frac{1}{\sqrt{\pi \text{det}(\sigma)}}\exp\left[-\frac{1}{2}x_i\left(\sigma^{-1}\right)_{ij}x_j\right] $. The displacement decays at a rate 2κ and will quickly reach 0. Thus, this function is entirely caracterised by the covariance matrix, which is described by the following equation: \begin{align} \partial_t\sigma = B\sigma+\sigma B^T- 2\kappa(\sigma - \sigma^L) \end{align} where \begin{align*} B =\begin{bmatrix} 0 & \omega_0 \\ \omega_0(Xg^2-1) & 0 \end{bmatrix} \end{align*}

In order to from F-P equation to covariance matrix equation, i insert Gaussian ansatz into F-P equation, ie \begin{align} \begin{split} \frac{\partial }{\partial t}(x,p)\frac{1}{\sqrt{\pi \text{det}(\sigma)}}\exp\left[-\frac{1}{2}x_i\left(\sigma^{-1}\right)_{ij}x_j\right]= &-\omega_0 \left(p\frac{\partial }{\partial x} - x\frac{\partial }{\partial p}\right)\frac{1}{\sqrt{\pi \text{det}(\sigma)}}\exp\left[-\frac{1}{2}x_i\left(\sigma^{-1}\right)_{ij}x_j\right]\\ &- \omega_0 X g^2 x\frac{\partial }{\partial p}\frac{1}{\sqrt{\pi \text{det}(\sigma)}}\exp\left[-\frac{1}{2}x_i\left(\sigma^{-1}\right)_{ij}x_j\right]\\ &+ 2 \kappa \frac{1}{\sqrt{\pi \text{det}(\sigma)}}\exp\left[-\frac{1}{2}x_i\left(\sigma^{-1}\right)_{ij}x_j\right]\\ &+ \kappa \left(x \frac{\partial }{\partial x} + p \frac{\partial }{\partial p}\right)\frac{1}{\sqrt{\pi \text{det}(\sigma)}}\exp\left[-\frac{1}{2}x_i\left(\sigma^{-1}\right)_{ij}x_j\right]\\ & + \frac12\kappa\left(\frac{\partial }{\partial x\partial x} + \frac{\partial }{\partial p\partial p} \right)\frac{1}{\sqrt{\pi \text{det}(\sigma)}}\exp\left[-\frac{1}{2}x_i\left(\sigma^{-1}\right)_{ij}x_j\right]\\ &+ \frac12 Xg^2\frac{\Gamma \omega_0}{\Omega}\frac{\partial }{\partial p\partial p}\frac{1}{\sqrt{\pi \text{det}(\sigma)}}\exp\left[-\frac{1}{2}x_i\left(\sigma^{-1}\right)_{ij}x_j\right] \end{split} \end{align} Then, simplify above equation, have \begin{align} \frac{\partial }{\partial t}(x,p)\frac{1}{\sqrt{\pi \text{det}(\sigma)}}\exp\left[-\frac{1}{2}x_i\left(\sigma^{-1}\right)_{ij}x_j\right] = &(-\omega_0 p + \kappa x)\frac{\partial }{\partial x}\frac{1}{\sqrt{\pi \text{det}(\sigma)}}\exp\left[-\frac{1}{2}x_i\left(\sigma^{-1}\right)_{ij}x_j\right]\\ &+(\omega_0 x - \omega_0 X g^2 x + \kappa p)\frac{\partial }{\partial p}\frac{1}{\sqrt{\pi \text{det}(\sigma)}}e^{-\frac{1}{2}x_i(\sigma^{-1})_{ij}x_j}\\ &+ 2 \kappa \frac{1}{\sqrt{\pi \text{det}(\sigma)}}\exp\left[-\frac{1}{2}x_i\left(\sigma^{-1}\right)_{ij}x_j\right]\\ & + \frac12\kappa\frac{\partial }{\partial x\partial x}\frac{1}{\sqrt{\pi \text{det}(\sigma)}}\exp\left[-\frac{1}{2}x_i\left(\sigma^{-1}\right)_{ij}x_j\right]\\ & + (\frac12\kappa + \frac12 Xg^2\frac{\Gamma \omega_0}{\Omega} ) \frac{\partial }{\partial p\partial p}\frac{1}{\sqrt{\pi \text{det}(\sigma)}}\exp\left[-\frac{1}{2}x_i\left(\sigma^{-1}\right)_{ij}x_j\right] \end{align} Next, evaluate the individual terms in the expression, have \begin{align} &\frac{\partial }{\partial t}(x)\frac{1}{\sqrt{\pi \text{det}(\sigma)}}\exp\left[-\frac{1}{2}x_i\left(\sigma^{-1}\right)_{ij}x_j\right]\\ = &\frac{\exp\left[-\frac{1}{2}x_i\left(\sigma^{-1}\right)_{ij}x_j\right] \left( -\frac{1}{2}x_i'(t)(\sigma(t)^{-1})_{ij} x_j (t) -\frac{1}{2}x_i(t) (\sigma(t)^{-1})_{ij} x_j'(t) -\frac{1}{2}x_i(t) (\sigma(t)^{-1})_{ij}'x_j (t) \right) }{\sqrt{\pi }\sqrt{\sigma_{11}(t) \sigma_{22}(t)-\sigma_{12}(t) \sigma_{21}(t)}}\\ &-\frac{\exp\left[-\frac{1}{2}x_i\left(\sigma^{-1}\right)_{ij}x_j\right] \left( \sigma_{11}'(t) \sigma_{22}(t) + \sigma _{11}(t) \sigma_{22}'(t) - \sigma_{12}'(t) \sigma_{21}(t) - \sigma _{12}(t) \sigma_{21}'(t) \right) }{2 \sqrt{\pi }\left(\sigma_{11}(t) \sigma_{22}(t)-\sigma_{12}(t) \sigma_{21}(t)\right)^{3/2}} \end{align} and \begin{align} &\frac{\partial }{\partial t}(x)\frac{1}{\sqrt{\pi \text{det}(\sigma)}}e^{-\frac{1}{2}x_i(\sigma^{-1})_{ij}x_j}\\ = &\frac{e^{-\frac{1}{2} x_i(t)(\sigma(t)^{-1})_{ij} x_j(t)} \left( -\frac{1}{2}x_i'(t)(\sigma(t)^{-1})_{ij} x_j (t) -\frac{1}{2}x_i(t) (\sigma(t)^{-1})_{ij} x_j'(t) -\frac{1}{2}x_i(t) (\sigma(t)^{-1})_{ij}'x_j (t) \right) }{\sqrt{\pi }\sqrt{\sigma_{11}(t) \sigma_{22}(t)-\sigma_{12}(t) \sigma_{21}(t)}}\\ &-\frac{e^{-\frac{1}{2} x_i(t)(\sigma(t)^{-1})_{ij} x_j(t)} \left( \sigma_{11}'(t) \sigma_{22}(t) + \sigma _{11}(t) \sigma_{22}'(t) - \sigma_{12}'(t) \sigma_{21}(t) - \sigma _{12}(t) \sigma_{21}'(t) \right) }{2 \sqrt{\pi }\left(\sigma_{11}(t) \sigma_{22}(t)-\sigma_{12}(t) \sigma_{21}(t)\right)^{3/2}} \end{align} and \begin{align} \frac{\partial }{\partial p}\frac{1}{\sqrt{\pi \text{det}(\sigma)}}e^{-\frac{1}{2}x_i(\sigma^{-1})_{ij}x_j} = &\frac{e^{-\frac{1}{2} x_i(\sigma^{-1})_{ij} x_j} \left( -\frac{1}{2}x_i'(\sigma^{-1})_{ij} x_j -\frac{1}{2}x_i (\sigma^{-1})_{ij} x_j' -\frac{1}{2}x_i (\sigma^{-1})_{ij}'x_j \right) }{\sqrt{\pi }\sqrt{\sigma_{11} \sigma_{22}-\sigma_{12} \sigma_{21}}}\\ &-\frac{e^{-\frac{1}{2} x_i(\sigma^{-1})_{ij} x_j} \left( \sigma_{11}' \sigma_{22} + \sigma _{11} \sigma_{22}' - \sigma_{12}' \sigma_{21} - \sigma _{12} \sigma_{21}' \right) }{2 \sqrt{\pi }\left(\sigma_{11} \sigma_{22}-\sigma_{12} \sigma_{21}\right)^{3/2}} \end{align} and \begin{align} &\frac{\partial }{\partial x \partial x}\frac{1}{\sqrt{\pi \text{det}(\sigma)}}e^{-\frac{1}{2}x_i(\sigma^{-1})_{ij}x_j}\\ = &\frac{e^{-\frac12 x_i\sigma^{-1}_{ij} x_j} \left(-\frac12 x_i (\sigma^{-1}_{ij})'x_j - \frac12 x_i'\sigma^{-1}_{ij} x_j -\frac12 x_i \sigma^{-1}_{ij} x_j'\right)^2}{\sqrt{\pi }\sqrt{\sigma_{11}\sigma_{22}-\sigma_{12}\sigma_{21}}}\\ &-\frac{-\frac12e^{-\frac12 x_i \sigma^{-1}_{ij} x_j} \left(\sigma_{11}'\sigma_{22} + \sigma _{11}\sigma_{22}' - \sigma_{12}'\sigma _{21} - \sigma_{12}\sigma_{21}'\right) \left( x_i (\sigma^{-1}_{ij})' x_j - x_i' \sigma^{-1}_{ij} x_j - x_i\sigma^{-1}_{ij} x_j'\right)}{\sqrt{\pi }\left(\sigma_{11}\sigma_{22} - \sigma_{12}\sigma_{21}\right)^{3/2}}\\ & + \frac{e^{-\frac{1}{2} x_i \sigma^{-1}_{ij} x_j} \left( -\frac{1}{2} x_i''\sigma^{-1}_{ij} x_j - \frac{1}{2} x_i (\sigma^{-1}_{ij})'' x_j - \frac{1}{2} x_i\sigma^{-1}_{ij} x_j'' - x_i'(\sigma^{-1}_{ij})'x_j - x_i(\sigma^{-1}_{ij})' x_j' - x_i'\sigma^{-1}_{ij} x_j'\right)}{\sqrt{\pi } \sqrt{\sigma _{11}\sigma_{22} - \sigma_{12} \sigma _{21}}}\\ &-\frac{\left(-2 \sigma _{12}' \sigma_{21}' + 2 \sigma _{11}' \sigma _{22}' + \sigma_{22}\sigma_{11}'' - \sigma_{21} \sigma_{12}'' - \sigma_{12}\sigma_{21}'' + \sigma _{11} \sigma _{22}''\right) e^{-\frac{1}{2} x_i\sigma^{-1}_{ij} x_j}}{2 \sqrt{\pi } \left(\sigma_{11}\sigma_{22} - \sigma_{12} \sigma_{21}\right)^{3/2}}\\ &+\frac{3 \left(\sigma_{22}\sigma_{11}' - \sigma_{21} \sigma_{12}' -\sigma _{12}\sigma _{21}'+\sigma _{11}\sigma _{22}'\right)^2 e^{-\frac{1}{2} x_i\sigma^{-1}_{ij} x_j}}{4 \sqrt{\pi } \left(\sigma _{11} \sigma _{22}-\sigma _{12} \sigma _{21}\right)^{5/2}} \end{align} et al. where we set \begin{align*} \sigma =\begin{bmatrix} \sigma_{11} & \sigma_{12} \\ \sigma_{21} & \sigma_{22}. \end{bmatrix} \end{align*} use $\partial_t x=0, \partial_x x=1,\partial_x (\sigma^{-1})_{11}=0,\partial_p x=0$ and simpliy above expression, have \begin{align} \begin{split} \frac{\partial}{\partial x}\frac{1}{\sqrt{\pi \text{det}(\sigma)}}\exp\left[-\frac{1}{2}x_1(\sigma^{-1})_{11}x_1\right] = &\frac{-\exp\left[-\frac{1}{2} x_1(\sigma^{-1})_{11} x_1\right] x_1\sigma^{-1}_{11} }{\sqrt{\pi }\sqrt{\sigma_{11} \sigma_{22}-\sigma_{12} \sigma_{21}}}\\ \frac{\partial}{\partial x}\frac{1}{\sqrt{\pi \text{det}(\sigma)}}\exp\left[-\frac{1}{2}x_2(\sigma^{-1})_{22}x_2\right] =&0\\ \frac{\partial}{\partial x}\frac{1}{\sqrt{\pi \text{det}(\sigma)}}\exp\left[-\frac{1}{2}x_1(\sigma^{-1})_{12}x_2\right] =&\frac{ -\frac{1}{2}\exp\left[-\frac{1}{2} x_1(\sigma^{-1})_{12} x_2\right](\sigma^{-1})_{12} x_2 }{\sqrt{\pi }\sqrt{\sigma_{11} \sigma_{22}-\sigma_{12} \sigma_{21}}}\\ \frac{\partial}{\partial p}\frac{1}{\sqrt{\pi \text{det}(\sigma)}}\exp\left[-\frac{1}{2}x_1(\sigma^{-1})_{11}x_1\right]=&0\\ \frac{\partial}{\partial p}\frac{1}{\sqrt{\pi \text{det}(\sigma)}}\exp\left[-\frac{1}{2}x_2(\sigma^{-1})_{22}x_2\right]=& \frac{-\exp\left[-\frac{1}{2} x_2(\sigma^{-1})_{22} x_2\right](\sigma^{-1})_{22} x_2 }{\sqrt{\pi }\sqrt{\sigma_{11} \sigma_{22}-\sigma_{12} \sigma_{21}}}\\ \frac{\partial}{\partial p}\frac{1}{\sqrt{\pi \text{det}(\sigma)}}\exp\left[-\frac{1}{2}x_1(\sigma^{-1})_{12}x_2\right]=& \frac{-\frac{1}{2}\exp\left[-\frac{1}{2} x_1(\sigma^{-1})_{12} x_2\right]x_1 (\sigma^{-1})_{12}}{\sqrt{\pi }\sqrt{\sigma_{11} \sigma_{22}-\sigma_{12} \sigma_{21}}}\\ \frac{\partial}{\partial x \partial x}\frac{1}{\sqrt{\pi \text{det}(\sigma)}}\exp\left[-\frac{1}{2}x_1(\sigma^{-1})_{11}x_1\right] = &\frac{\exp\left[-\frac12 x_1\sigma^{-1}_{11} x_1\right] \left(- \sigma^{-1}_{11} x_1 \right)^2}{\sqrt{\pi }\sqrt{\sigma_{11}\sigma_{22}-\sigma_{12}\sigma_{21}}} + \frac{\exp\left[-\frac{1}{2} x_1 \sigma^{-1}_{11} x_1\right] \left( -\sigma^{-1}_{11}\right)}{\sqrt{\pi } \sqrt{\sigma _{11}\sigma_{22} - \sigma_{12} \sigma _{21}}}\\ \frac{\partial}{\partial x \partial x}\frac{1}{\sqrt{\pi \text{det}(\sigma)}}\exp\left[-\frac{1}{2}x_2(\sigma^{-1})_{22}x_2\right] = & 0\\ \frac{\partial}{\partial x \partial x}\frac{1}{\sqrt{\pi \text{det}(\sigma)}}\exp\left[-\frac{1}{2}x_1(\sigma^{-1})_{12}x_2\right] = &\frac{\exp\left[-\frac12 x_1\sigma^{-1}_{12}x_2\right] \left(- \frac12 \sigma^{-1}_{12}x_2\right)^2}{\sqrt{\pi }\sqrt{\sigma_{11}\sigma_{22}-\sigma_{12}\sigma_{21}}}\\ \frac{\partial}{\partial p \partial p}\frac{1}{\sqrt{\pi \text{det}(\sigma)}}\exp\left[-\frac{1}{2}x_1(\sigma^{-1})_{11}x_1\right]= & 0\\ \frac{\partial}{\partial p \partial p}\frac{1}{\sqrt{\pi \text{det}(\sigma)}}\exp\left[-\frac{1}{2}x_2(\sigma^{-1})_{22}x_2\right]= & \frac{\exp\left[-\frac12 x_2\sigma^{-1}_{22} x_2\right] \left( - \sigma^{-1}_{22} x_2 \right)^2}{\sqrt{\pi }\sqrt{\sigma_{11}\sigma_{22}-\sigma_{12}\sigma_{21}}} + \frac{\exp\left[-\frac{1}{2} x_2 \sigma^{-1}_{22} x_2\right] \left( -\sigma^{-1}_{22}\right)}{\sqrt{\pi } \sqrt{\sigma _{11}\sigma_{22} - \sigma_{12} \sigma _{21}}}\\ \frac{\partial}{\partial p \partial p}\frac{1}{\sqrt{\pi \text{det}(\sigma)}}\exp\left[-\frac{1}{2}x_1(\sigma^{-1})_{12}x_2\right]= & \frac{\exp\left[-\frac12 x_1\sigma^{-1}_{12}x_2\right] \left(-\frac12 x_1 \sigma^{-1}_{12}\right)^2}{\sqrt{\pi }\sqrt{\sigma_{11}\sigma_{22}-\sigma_{12}\sigma_{21}}} \end{split} \end{align} Last, insert them back into the F-P equation and spilt equation into $\partial_t \sigma_{11},\partial_t \sigma_{22},\partial_t \sigma_{12}$, we have \begin{align} \begin{split} &\frac{\partial }{\partial t}(x,p)\frac{1}{\sqrt{\pi \text{det}(\sigma)}}e^{-\frac{1}{2}x_1(\sigma^{-1})_{11}x_1} = (-\omega_0 p + \kappa x)\frac{\partial }{\partial x}\frac{1}{\sqrt{\pi \text{det}(\sigma)}}e^{-\frac{1}{2}x_1(\sigma^{-1})_{11}x_1}\\ &+(\omega_0 x - \omega_0 X g^2 x + \kappa p)\frac{\partial }{\partial p}\frac{1}{\sqrt{\pi \text{det}(\sigma)}}e^{-\frac{1}{2}x_1(\sigma^{-1})_{11}x_1}\\ &+ 2 \kappa \frac{1}{\sqrt{\pi \text{det}(\sigma)}}e^{-\frac{1}{2}x_1(\sigma^{-1})_{11}x_1}\\ & + \frac12\kappa\frac{\partial }{\partial x\partial x}\frac{1}{\sqrt{\pi \text{det}(\sigma)}}e^{-\frac{1}{2}x_1(\sigma^{-1})_{11}x_1}\\ & + (\frac12\kappa + \frac12 Xg^2\frac{\Gamma \omega_0}{\Omega} ) \frac{\partial }{\partial p\partial p}\frac{1}{\sqrt{\pi \text{det}(\sigma)}}e^{-\frac{1}{2}x_1(\sigma^{-1})_{11}x_1}\\ \Rightarrow&\\ &\frac{e^{-\frac{1}{2} x_i(t)(\sigma(t)^{-1})_{ij} x_j(t)} \left( -\frac{1}{2}x_1(t) (\sigma(t)^{-1})_{11}'x_1 (t) \right) }{\sqrt{\pi }\sqrt{\sigma_{11}(t) \sigma_{22}(t)-\sigma_{12}(t) \sigma_{21}(t)}}\\ &-\frac{e^{-\frac{1}{2} x_1(t)(\sigma(t)^{-1})_{11} x_1(t)} \left( \sigma_{11}'(t) \sigma_{22}(t) + \sigma _{11}(t) \sigma_{22}'(t) - \sigma_{12}'(t) \sigma_{21}(t) - \sigma _{12}(t) \sigma_{21}'(t) \right) }{2 \sqrt{\pi }\left(\sigma_{11}(t) \sigma_{22}(t)-\sigma_{12}(t) \sigma_{21}(t)\right)^{3/2}}\\ = &(-\omega_0 p + \kappa x)\frac{-e^{-\frac{1}{2} x_1(\sigma^{-1})_{11} x_1} x_1\sigma^{-1}_{11} }{\sqrt{\pi }\sqrt{\sigma_{11} \sigma_{22}-\sigma_{12} \sigma_{21}}}\\ &+(\omega_0 x - \omega_0 X g^2 x + \kappa p)0\\ &+ 2 \kappa \frac{1}{\sqrt{\pi }\sqrt{\sigma_{11} \sigma_{22}-\sigma_{12} \sigma_{21}}}e^{-\frac{1}{2}x_1(\sigma^{-1})_{11}x_1}\\ & + \frac12\kappa\left(\frac{e^{-\frac12 x_1\sigma^{-1}_{11} x_1} \left(- \sigma^{-1}_{11} x_1 \right)^2}{\sqrt{\pi }\sqrt{\sigma_{11}\sigma_{22}-\sigma_{12}\sigma_{21}}} + \frac{e^{-\frac{1}{2} x_1 \sigma^{-1}_{11} x_1} \left( -\sigma^{-1}_{11}\right)}{\sqrt{\pi } \sqrt{\sigma _{11}\sigma_{22} - \sigma_{12} \sigma _{21}}}\right) \end{split} \end{align} simplify, have \begin{align} \partial_t\sigma(t) = -2(-\omega_0 p + \kappa x)/x_1\sigma_{11} + 4 \kappa\sigma_{11}^2/x_1^2 + \kappa -\kappa /x_1\sigma_{11} \end{align} we obtained the result of $\partial_t \sigma_{11}$ is different the result from covariance matrix equation \begin{align} \begin{split} \partial_t \sigma_{11} =& \omega_0\sigma_{21} + \omega_0\sigma_{12} - 2\kappa(\sigma_{11} - \frac12)\\ \partial_t \sigma_{12} =& \omega_0\sigma_{22} + \omega_0(Xg^2-1)\sigma_{11} - 2\kappa\sigma_{12}\\ \partial_t \sigma_{21} =& \omega_0\sigma_{22} + \omega_0(Xg^2-1)\sigma_{11} - 2\kappa\sigma_{21}\\ \partial_t \sigma_{22} =& \omega_0(Xg^2-1)\sigma_{12} + \omega_0(Xg^2-1)\sigma_{21} - 2\kappa(\sigma_{22} -\frac12 - \frac12Xg^2\frac{\Gamma\omega_0}{\Omega\kappa}) \end{split} \end{align}

I don't know if I miscalculated or missed something, or if I was thinking in the wrong way. I have no idea for this equation, it's been confusing me for a long time.

Any little help would be good for me. Any hint? Thank you very much!

New contributor
gang liu is a new contributor to this site. Take care in asking for clarification, commenting, and answering. Check out our Code of Conduct.
$\endgroup$

0

Your Answer

gang liu is a new contributor. Be nice, and check out our Code of Conduct.

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.