Modeling pulse propagation in fiber optical parametric amplifier by a long short-term memory network

The ultrashort pulse evolution in fiber optical parametric amplifier (FOPA) systems is a highly complex nonlinear dynamic. Here, the long short-term memory network (LSTM) is applied to predicting the pulse propagation of signal waves in the FOPA system, rather than calculating the nonlinear Schrödinger equation (NLSE) numerically. The LSTM network only requires the initial signal power waveform, and the signal pulse propagation is realized in the feedback loop. Besides, we focus on predicting the evolution of signal pulse by varying the initial signal power and the initial signal pulse width to test the performance of our proposed method. The predictions of the LSTM network show great consistency with the NLSE results in each propagation stage of two cases, which reveal that the signal pulse evolution in FOPA is achieved precisely. The realization reduces the computational burden and further gives an effective way to optimize and guide the experiment and application in the FOPA systems.