Solve a nonlinear matrix equation

Join the discussion
Ask a follow-up here, or get your own question answered by working scientists, mathematicians and engineers — people, not an autocomplete.
Real named experts · corrections over time · the nuance an AI answer skips
12 replies · 2K views
rehan_eme
Messages
1
Reaction score
0
TL;DR
Solve a nonlinear matrix equation
Hi all,

I want to know if a second solution exists for the following math equation:

Ce^{At} ρ_p+(CA)^{−1} (e^{At}−I)B=0

Where C, ρ_p, A and B are constant matrices, 't' is scalar variable. I know that atleast one solution i.e. 〖t=θ〗_1 exists, but I want a method to determine if there is another θ_0<θ_1 that is also a (second) solution. Any anlaytical way of determining that is what I am looking for.
 
Physics news on Phys.org
Mayhem said:
Doesn't this show up in QM quite often? I remember seeing something to the effect of ##e^{\mathbf{X}}## before.
The exponentiation comes into play whenever we want to solve a differential equation when we turn something linear into something curved, the transition from a Lie algebra to a Lie group. However, the OP didn't provide any such context, no differential equation, no vector field. It is not even clear whether ##e^{At}## converges. It looks like a flow, but it would have been nice to know for sure.
 
Vanadium 50 said:
I have not seen it in QM, or for that matter, ever. Raising a scalar to the identity matrix returns what? (Apparently, a disgonal matrix with the scalar on the diagonal,,,guess it has to be something)
Here is what I think it should have been:

rehan_eme said:
TL;DR Summary: Solve a nonlinear matrix equation

Hi all,

I want to know if a second solution exists for the following math equation:

##Ce^{At} ρ_p + (CA)^{−1} (e^{At}−I) B=0## or ##C\exp(At) ρ_p + (CA)^{−1} (\exp(At) −I ) B=0##

Where ##C, ρ_p, A,## and ##B## are constant matrices, ##t## is a scalar variable. I know that at least one solution i.e. ##t=\theta_1## exists, but I want a method to determine if there is another ##\theta_0<\theta_1## that is also a (second) solution. Any analytical way of determining that is what I am looking for.
E.g. it could be that ##A\in \mathfrak{su}(2)## and ##e^{At}## is a flow in ##\operatorname{SU}(2).##
 
Vanadium 50 said:
Raising a scalar to the identity matrix returns what? (Apparently, a disgonal matrix with the scalar on the diagonal,,,guess it has to be something)
This is something that appears in all but the most elementary linear algebra textbooks. As a Maclaurin expansion ##e^A = I + A + \frac {A^2}{2!} + \frac {A^3}{3!} + \dots + \frac {A^n}{n!} + \dots##.

##e^I = I + I + \frac {I^2}{2!} + \frac {I^3}{3!} + \dots + \frac {I^n}{n!} + \dots##
##= I(2 + 1/2 + 1/6 + 1/24 + \dots) = eI##
 
Reply
  • Like
Likes   Reactions: DaveE
Vanadium 50 said:
I have not seen it in QM, or for that matter, ever.
That's surprising to me. Maybe I'm showing my age here, but back in the day, quantum mechanics textbooks often exponentiated the time-independent, Hermetian Hamiltonian ##H## (a differential operator or a matrix, depending on the representation) of a quantum system to define the unitary operator ##T## of time evolution (e.g., see the page from Merzbacher below). Maybe the QM texts you've studied no longer use that approach?

Merzbacher QM.jpg
 

Attachments

  • Merzbacher QM.jpg
    Merzbacher QM.jpg
    68.9 KB · Views: 146
Reply
  • Like
Likes   Reactions: weirdoguy