A data-driven approach to solving a 1D inverse scattering problem

Tristan van Leeuwen,Andreas Tataris

2023-04-06

Abstract:In this paper, we extend the ROM-based approach for inverse scattering with Neumann boundary conditions, introduced by Druskin at. al. (Inverse Problems 37, 2021), to the 1D Schr{ö}dinger equation with impedance (Robin) boundary conditions. We also propose a novel data-assimilation (DA) inversion method based on the ROM approach, thereby avoiding the need for a Lanczos-orthogonalization (LO) step. Furthermore, we present a detailed numerical study and comparison of the accuracy and stability of the DA and LO methods.

Numerical Analysis,Mathematical Physics

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the problem of inverting the scattering potential from the boundary data in the one - dimensional Schrödinger equation. Specifically, the goal of the paper is to expand the technology based on the Reduced Order Model (ROM) through a data - driven method to solve the one - dimensional inverse scattering problem with impedance boundary conditions. Traditional methods usually require a Lanczos orthogonalization step, while this paper proposes a new Data - Assimilation (DA) method, which avoids this complicated step, and conducts a detailed numerical study and comparison between the new method and the traditional method in terms of accuracy and stability.

### The main contributions of the paper include:
1. **Expansion of the ROM method**: Expand the ROM method originally applied to the one - dimensional diffusion problem (Neumann boundary condition) to the one - dimensional Schrödinger equation (impedance boundary condition).
2. **Data - assimilation method**: Propose a new data - assimilation method for estimating the state from the boundary data, avoiding the traditional Lanczos orthogonalization step.
3. **Numerical experiments**: Compare the performance of the data - assimilation method and the Lanczos orthogonalization method under different noise levels through numerical experiments, and verify the accuracy and stability of the new method.

### The specific content of the paper:
- **Forward problem**: Describe the one - dimensional Schrödinger equation and its impedance boundary conditions, and give the form of the measurement data.
- **Reduced - order model**: Explain in detail how to calculate the ROM matrices from the boundary data, and propose a state - estimation method based on these matrices.
- **Inverse problem**: Divided into two steps: first, estimate the state from the data, and then recover the scattering potential from the estimated state.
- **State estimation**: Discuss two methods: Lanczos orthogonalization and data - assimilation.
- **Scattering - potential estimation**: Derive the Lippmann - Schwinger - type equation using the weak form of the differential equation, and solve the scattering potential by the least - squares method.
- **Numerical results**: Show the performance comparison of the two methods under noise - free and noisy data, indicating that the data - assimilation method provides more stable results at moderate noise levels.

### Conclusion:
The paper successfully expands the ROM method to the one - dimensional Schrödinger equation with impedance boundary conditions, and proposes a new data - assimilation method, which shows better accuracy and stability in state estimation and scattering - potential recovery. These results lay the foundation for further applying the ROM method to frequency - domain wave problems (such as the Helmholtz equation) and two - dimensional/three - dimensional problems.