Solving the 2D Advection-Diffusion Equation using Fixed-Depth Symbolic Regression and Symbolic Differentiation without Expression Trees

Edward Finkelstein

2024-10-18

Abstract:This paper presents a novel method for solving the 2D advection-diffusion equation using fixed-depth symbolic regression and symbolic differentiation without expression trees. The method is applied to two cases with distinct initial and boundary conditions, demonstrating its accuracy and ability to find approximate solutions efficiently. This framework offers a promising, scalable solution for finding approximate solutions to differential equations, with the potential for future improvements in computational performance and applicability to more complex systems involving vector-valued objectives.

Computation,Machine Learning,Numerical Analysis

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to solve the 2D advection - diffusion equation through fixed - depth symbolic regression and symbolic differentiation without expression trees. Specifically, the author proposes a novel method, aiming to efficiently find the approximate solution of this partial differential equation and verify its accuracy and efficiency under different initial and boundary conditions.

### Specific problem description

1. **2D advection - diffusion equation**:
The paper focuses on the 2D advection - diffusion equation in the following form:
\[
\frac{\partial T}{\partial t}+\nabla\cdot(\vec{u}T)=\kappa\nabla^{2}T
\]
where:
- \(T(x, y, t)\) is the temperature field or concentration field.
- \(\vec{u}=(u_{x}, u_{y})\) is the velocity field.
- \(\kappa\) is the diffusion coefficient.

2. **Innovation points of the method**:
- **Fixed - depth symbolic regression**: By limiting the depth of expressions, the search space is reduced and the computational efficiency is improved.
- **Symbolic differentiation without expression trees**: Using prefix or suffix notation (Polish notation or Reverse Polish notation), the memory overhead and performance bottlenecks brought by the traditional expression tree structure are avoided.

3. **Objectives**:
- To find an approximate analytical solution \(T(x, y, t)\) that can satisfy the given initial and boundary conditions.
- To verify the effectiveness and accuracy of this method under different conditions.

4. **Application scenarios**:
This method can be applied to the modeling of physical phenomena such as heat conduction, and the diffusion of gases and fluids in various media.

### Summary

The main purpose of the paper is to present a new computational framework, which uses fixed - depth symbolic regression and symbolic differentiation techniques to efficiently solve the 2D advection - diffusion equation without relying on expression trees. This method not only improves the computational efficiency, but also provides potential application prospects for solving other complex partial differential equations.