Abstract:In this paper, we present a new multiscale domain decomposition algorithm for computing solutions of static Eikonal equations. The new method is an iterative two-scale method that uses a parareal-like update scheme in combination with standard Eikonal solvers. The purpose of the two scales is to accelerate convergence and maintain accuracy. We adapt a weighted version of the parareal method for stability, and the optimal weights are studied via a model problem. Numerical examples are given to demonstrate the method.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is: how to efficiently solve the boundary value problem of the static Eikonal equation with multi - scale features. Specifically, the author proposes a new multi - scale domain decomposition algorithm to accelerate convergence and maintain the accuracy of the numerical solution.
### Problem Background
The Eikonal equation has wide applications in optimal control, path planning, seismology, geometric optics, etc. This equation is a fully nonlinear Hamilton - Jacobi equation, and usually there is no classical solution, so it is necessary to solve the unique viscosity solution. The paper focuses on the static Eikonal equation in the following form:
\[
|\nabla u(x)| = r_\epsilon(x), \quad x \in \Omega \subset \mathbb{R}^d \tag{1.1}
\]
\[
u(x) = g(x), \quad x \in \Gamma \subset \partial \Omega \tag{1.2}
\]
where \( r_\epsilon(x) = r_0(x) + a_\epsilon(x) \), \( r_0 \) is a smoothing function, and \( a_\epsilon \) describes the multi - scale features.
### Limitations of Existing Methods
The existing serial algorithms (such as the Fast Sweeping Method (FSM) and the Fast Marching Method (FMM)) have limitations when dealing with large - scale discrete systems. For the Eikonal equation with multi - scale features, in order to accurately capture the fine - scale features, very fine grid discretization is required, which leads to a large number of coupled nonlinear equations, with high computational cost and slow speed.
### The Method Proposed in the Paper
To solve the above problems, the author proposes a new multi - scale domain decomposition algorithm. This algorithm combines the characteristics of the parareal method and the standard Eikonal equation solver, aiming to accelerate convergence and maintain accuracy. The specific steps include:
1. **Initialization**: Use the Fast Sweeping Method (FSM) to initialize the solution on the coarse grid and record the wind direction (i.e., the direction of the minimum value of the neighboring grid points).
2. **Update Sub - domain Boundary Conditions**: Set the boundary conditions of the sub - domains according to the coarse - grid solution to ensure the correct propagation of information.
3. **Solve the Fine - grid Solution in Parallel**: Solve the Eikonal equation in parallel within each sub - domain and record the wind direction on the fine grid.
4. **Coarse - grid Update**: Use weighted correction to update the coarse - grid solution, ensure the correct causality, and perform causal sweeping to correct possible incorrect causal relationships.
Through this method, the author hopes to significantly improve the solution speed while maintaining accuracy, especially for the Eikonal equation with multi - scale features.
### Summary
The main goal of this paper is to develop an efficient multi - scale domain decomposition algorithm for solving the boundary value problem of the static Eikonal equation with multi - scale features. By combining the parareal method and the standard Eikonal equation solver, this algorithm aims to accelerate convergence and maintain the accuracy of the numerical solution.