Revisiting the dynamics of Bose-Einstein condensates in a double well by deep learning with a hybrid network

TL;DR

This paper introduces a hybrid deep learning approach combining LSTM and ResNet to efficiently simulate the complex, oscillatory dynamics of Bose-Einstein condensates in a double-well potential, outperforming single-network methods.

Contribution

The authors develop a novel hybrid neural network method that accurately predicts strongly-oscillating quantum system dynamics, overcoming limitations of traditional single-network approaches.

Findings

High-fidelity prediction of BEC dynamics achieved

Efficient pre-learning process demonstrated

Applicable to complex systems with multiple oscillation frequencies

Abstract

Deep learning, accounting for the use of an elaborate neural network, has recently been developed as an efficient and powerful tool to solve diverse problems in physics and other sciences. In the present work, we propose a novel learning method based on a hybrid network integrating two different kinds of neural networks: Long Short-Term Memory(LSTM) and Deep Residual Network(ResNet), in order to overcome the difficulty met in numerically simulating strongly-oscillating dynamical evolutions of physical systems. By taking the dynamics of Bose-Einstein condensates in a double-well potential as an example, we show that our new method makes a high efficient pre-learning and a high-fidelity prediction about the whole dynamics. This benefits from the advantage of the combination of the LSTM and the ResNet and is impossibly achieved by a single network in the case of direct learning. Our method can be applied for simulating complex cooperative dynamics in a system with fast multiple-frequency oscillations with the aid of auxiliary spectrum analysis.