Predicting the evolution of the mid-infrared supercontinuum in the near-zero dispersion region under weak triggering using LSTM neural networks

Abstract

To enable active control of rogue waves after pumping a pulse into an optical fiber, adding a weak trigger is necessary. However, this process typically involves numerous numerical simulations based on the nonlinear Schrodinger equation, which can be computationally intensive. Using neural networks to predict complex nonlinear propagation in optical fibers avoids the need for numerical solutions during research. We show a neural network using long short-term memory to predict the temporal and spectral evolutions of the mid-infrared supercontinuum generation in the near-zero dispersion region under weak triggering, simply by determining the input pulse strength in the time and frequency domains.