A feed-forward neural network as a nonlinear dynamics integrator for supercontinuum generation

TL;DR

This paper introduces a feed-forward neural network that efficiently models the nonlinear dynamics of supercontinuum generation in optical fibers, enabling faster and less resource-intensive simulations compared to traditional methods.

Contribution

The authors develop a novel feed-forward neural network model that accurately emulates the GNLSE for supercontinuum generation, offering improvements over recurrent neural networks in speed and memory usage.

Findings

The neural network accurately reproduces supercontinuum generation dynamics.

The feed-forward model trains faster and requires less memory than recurrent models.

The approach is adaptable to other physical systems involving nonlinear dynamics.

Abstract

The nonlinear propagation of ultrashort pulses in optical fiber depends sensitively on both input pulse and fiber parameters. As a result, optimizing propagation for specific applications generally requires time-consuming simulations based on sequential integration of the generalized nonlinear Schr\"odinger equation (GNLSE). Here, we train a feed-forward neural network to learn the differential propagation dynamics of the GNLSE, allowing emulation of direct numerical integration of fiber propagation, and particularly the highly complex case of supercontinuum generation. Comparison with a recurrent neural network shows that the feed-forward approach yields faster training and computation, and reduced memory requirements. The approach is generic and can be extended to other physical systems.