Deep Neural Network for Predicting Supercontinuum Broadening in Chalcogenide Photonic Quasi crystal Fiber

Generation SuperContinuum (SC) is a complex nonlinear process caused by chaotic and unstable behaviour, especially simulated by short duration pump pulses when the optical fiber work on the anomalous dispersion region. Understanding the spectral broadening behavior is difficult due to variations in fiber length and input pulse parameters. To address this challenge, we introduce a Deep Neural Network (DNN) to predict the SC spectrum broadening. Remarkably, the DNN provides accurate predictions across different lengths of Chalcogenide (ChG) Photonic quasi crystal Fiber (PQF), achieving a root mean square error (RMSE) of 1.3677 and an R-Squared value of 0.99. Additionally, this DNN significantly reduces the computational time compared to MATLAB and COMSOL Multiphysics software.