Predicting electromagnetically induced transparency based cold atomic engines using deep learning

• Published in: APL quantum Vol. 2; no. 2; pp. 026132 - 026132-11
• Main Authors: Sarmah, Manash Jyoti, Goswami, Himangshu Prabal
• Format: Journal Article
• Language: English
• Published: AIP Publishing LLC 01.06.2025
• ISSN: 2835-0103; 2835-0103
• DOI: 10.1063/5.0255830

We develop an artificial neural network model to predict quantum heat engines working within the experimentally realized framework of electromagnetically induced transparency. We specifically focus on Λ-type alkali-based cold atomic systems. This network allows us to analyze the performance of all the alkali atom-based engines. High performance engines are predicted and analyzed based on three figures of merit: output radiation temperature, work, and ergotropy. Contrary to traditional notion, the algorithm reveals the limitations of output radiation temperature as a standalone metric for enhanced engine performance. In high-output radiation temperature regime, a Cs-based engine with a higher output-temperature than a Rb-based engine is characterized by lower work and ergotropy. This is found to be true for different atomic engines with common predicted states in both high- and low-output radiation temperature regimes. In addition, the ergotropy is found to exhibit a saturating exponential dependency on the control Rabi frequency.