Abstract:In this work, we solve inverse problems of nonlinear Schrödinger equations that can be formulated as a learning process of a special convolutional neural network. Instead of attempting to approximate functions in the inverse problems, we embed a library as a low dimensional manifold in the network such that unknowns can be reduced to some scalars. The nonlinear Schrödinger equation (NLSE) is $i\frac{\partial \psi}{\partial t}-\beta\frac{\partial^2 \psi}{\partial x^2}+\gamma|\psi|^2\psi+V(x)\psi=0,$ where the wave function $\psi(x,t)$ is the solution to the forward problem and the potential $V(x)$ is the quantity of interest of the inverse problem. The main contributions of this work come from two aspects. First, we construct a special neural network directly from the Schrödinger equation, which is motivated by a splitting method. The physics behind the construction enhances explainability of the neural network. The other part is using a library-search algorithm to project the solution space of the inverse problem to a lower-dimensional space. The way to seek the solution in a reduced approximation space can be traced back to the compressed sensing theory. The motivation of this part is to alleviate the training burden in estimating functions. Instead, with a well-chosen library, one can greatly simplify the training process. A brief analysis is given, which focuses on well-possedness of some mentioned inverse problems and convergence of the neural network approximation. To show the effectiveness of the proposed method, we explore in some representative problems including simple equations and a couple equation. The results can well verify the theory part. In the future, we can further explore manifold learning to enhance the approximation effect of the library-search algorithm.

Linearized Inverse Schrödinger Potential Problem with Partial Data and Its Deep Neural Network Inversion

Linearized Inverse Schrödinger Potential Problem at a Large Wavenumber

Increasing Stability in the Linearized Inverse Schrödinger Potential Problem with Power Type Nonlinearities

Solving the inverse potential problem in the parabolic equation by the deep neural networks method

Partial data inverse problems for the nonlinear magnetic Schrödinger equation

A clever neural network in solving inverse problems of Schrödinger equation

Solving an Inverse Source Problem by Deep Neural Network Method with Convergence and Error Analysis

An inverse problem for a semi-linear wave equation: a numerical study

Increasing Stability of the First Order Linearized Inverse Schrödinger Potential Problem with Integer Power Type Nonlinearities

Recovering Simultaneously A Potential And A Point Source From Cauchy Data

Stability for the Inverse Potential Problem by Finite Measurements on the Boundary

Increasing stability of a linearized inverse boundary value problem for a nonlinear Schrödinger equation on transversally anisotropic manifolds

An Inverse Problem with Partial Neumann Data and $L^{n/2}$ Potentials

Soliton, Breather and Rogue Wave Solutions for Solving the Nonlinear Schrödinger Equation Using a Deep Learning Method with Physical Constraints

Lipschitz Stability of an Inverse Boundary Value Problem for a Schrödinger-Type Equation

A Linearized Boundary Control Method for the Acoustic Inverse Boundary Value Problem

Deep neural networks can stably solve high-dimensional, noisy, non-linear inverse problems

Stabilizing and Solving Inverse Problems using Data and Machine Learning

Deep neural networks learning forward and inverse problems of two-dimensional nonlinear wave equations with rational solitons

An Unsupervised Deep Learning Approach for the Wave Equation Inverse Problem

On the determination of the spatial-dependent potential coefficient in a linear pseudoparabolic equation