Abstract:We discuss through multiple numerical examples the accuracy and efficiency of a micro-macro acceleration method for stiff stochastic differential equations (SDEs) with a time-scale separation between the fast microscopic dynamics and the evolution of some slow macroscopic state variables. The algorithm interleaves a short simulation of the stiff SDE with extrapolation of the macroscopic state variables over a longer time interval. After extrapolation, we obtain the reconstructed microscopic state via a matching procedure: we compute the probability distribution that is consistent with the extrapolated state variables, while minimally altering the microscopic distribution that was available just before the extrapolation. In this work, we numerically study the accuracy and efficiency of micro-macro acceleration as a function of the extrapolation time step and as a function of the chosen macroscopic state variables. Additionally, we compare the effect of different hierarchies of macroscopic state variables. We illustrate that the method can take significantly larger time steps than the inner microscopic integrator, while simultaneously being more accurate than approximate macroscopic models.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is: for stiff stochastic differential equations (SDEs) with time - scale separation, how to improve the efficiency and accuracy of the micro - macro acceleration method. Specifically, the author studied the accuracy and efficiency of the micro - macro acceleration method under different extrapolation time steps and different selections of macro - state variables through numerical experiments, and compared it with the approximate macro - model.
### Background of the paper and problem description
Many stochastic differential equations (SDEs) used in scientific and engineering applications can be expressed as:
\[
dX_t = a(X_t, t)dt + b(X_t, t)dW_t
\]
where \(X_t\) is a diffusion process on the domain \(G\subset\mathbb{R}^d\), \(a(X_t, t)\in\mathbb{R}^d\) is the drift term, \(b(X_t, t)\in\mathbb{R}^{d\times n}\) is the diffusion term, and \(W_t\in\mathbb{R}^n\) represents an \(n\)-dimensional Brownian motion.
However, in practical applications, we are usually only interested in a few macro - state variables \(m(t)=(m_1(t),\ldots,m_L(t))\), which are defined as the expected values of certain functions of interest \(R(x)=(R_1(x),\ldots,R_L(x))\) on the distribution of \(X_t\):
\[
t\mapsto m(t)=\mathbb{E}[R(X_t)]
\]
When the SDE system is stiff, that is, the individual micro - paths change on a fast time - scale, while the evolution of the macro - state variables occurs on a longer time - scale, explicit time - discretization methods (such as the Euler - Maruyama scheme) are forced to adopt a small time step to ensure stability. Therefore, explicit methods are not suitable for simulation up to a large termination time \(T\).
### Micro - macro acceleration method
To overcome this problem, the author introduced a new micro - macro acceleration algorithm. This algorithm takes advantage of the stiffness characteristics of SDEs and introduces a second time step \(\Delta t\), which is much larger than the micro - time step \(\delta t\), for extrapolating the macro - state variables \((m_1(t),\ldots,m_L(t))\). A time step of one micro - macro acceleration includes four steps:
1. **Monte Carlo simulation**: Simulate the SDE using a small time step \(\delta t\).
2. **Restriction**: Estimate the macro - state variables from the set of micro - Monte Carlo particles.
3. **Extrapolation**: Extrapolate the macro - state variables over a larger time interval \(\Delta t\).
4. **Matching**: Generate a new micro - probability distribution that is consistent with the extrapolated macro - state while minimizing the deviation from the previous distribution.
### Main research content
This paper studied the influence of the following parameters on the accuracy of the micro - macro acceleration method through numerical experiments:
- The number \(L\) of macro - state variables
- The selected functions of interest \(R_l\) and their ability to capture the essential characteristics of the underlying micro - distribution
- The extrapolation step size \(\Delta t\)
The author also compared the selections of macro - state variables at different levels and showed that the micro - macro acceleration method can achieve higher accuracy under a given number of macro - state variables \(L\) and extrapolation step size \(\Delta t\), and is faster than a complete micro - simulation.
### Conclusion
Through the above research, the author proved that the micro - macro acceleration method can significantly improve computational efficiency while maintaining high accuracy, especially when dealing with stiff stochastic differential equations with time - scale separation.