iibewegung
Aug11-08, 08:44 PM
Hi,
Can anyone tell me an algorithm to generate the stationary Gaussian distribution R(t) with
\langle R(t) \rangle = 0 (zero mean)
\langle R(t) R(t')^{T} \rangle = A \delta(t-t') , A = 2 \gamma k_B T m (autocorrelation)
?
What I just wrote is from the Wikipedia article "Langevin dynamics"
and R(t) belongs to the simple Langevin equation
F(x) = m \ddot{x} = -\nabla U(x) - \gamma m \dot{x} + R(t)
Can anyone tell me an algorithm to generate the stationary Gaussian distribution R(t) with
\langle R(t) \rangle = 0 (zero mean)
\langle R(t) R(t')^{T} \rangle = A \delta(t-t') , A = 2 \gamma k_B T m (autocorrelation)
?
What I just wrote is from the Wikipedia article "Langevin dynamics"
and R(t) belongs to the simple Langevin equation
F(x) = m \ddot{x} = -\nabla U(x) - \gamma m \dot{x} + R(t)