What probability density is used in Brownian motion?

1 reply · 2K views · last reply
Earthland
Messages
27
Reaction score
0

Homework Statement



I have a free Brownian particle and its coordinate is given as a function of time:

Displayed equation: x(t) = x(t0) + v(t0)/γ [1 − e^(−γ(t−t0))] + (1/γ) ∫ from t0 to t of [1 − e^(−γ(t−t′))] ζ(t′)/m dt′.


And its first moment, or mean, is given as

Equation for the mean position of a damped Brownian particle: <x(t)> = x(t0) + (v(t0)/gamma)[1 - e^(-gamma(t - t0))].


But what kind of probability density was used to calculate this first moment?

Homework Equations



I know that the first moment is calculated by using probability density

<x> = ∫x*W(x)dx from -infinity to +infinity.

X itself is dependent on t, so is it W(x) or W(x,t) or W(t)? I'm little lost on this one.

The Attempt at a Solution



I did find some kind of density function in regards to Brownian motion

Gaussian heat-kernel density f_t(x) = (1/sqrt(2 pi t)) exp(-x^2/(2t)), a function of position x and time t, shown as an equation.


But I can't see how this would yield the correct result. And still, must I integrate over x or t?
 
Attachments
  • Equation for the position x(t) of a Langevin particle: x(t) = x(t0) + v(t0)/γ [1 − e^(−γ(t−t0))] + (1/γ) integral from t0 to t of [1 − e^(−γ(t−t′))] ζ(t′)/m dt′.
    upload_2015-12-4_13-57-39.png
    18.2 KB · Views: 501
Just take the expectation value of your first expression (expectation values are linear) keeping in mind that the only actually random variable in it is the driving force ##\zeta(t')##.