基于深度强化学习的量子奥托循环性能优化

doi:10.3969/j.issn.1007-5461.2025.01.007

量子电子学报 ›› 2025, Vol. 42 ›› Issue (1): 70-0.doi: 10.3969/j.issn.1007-5461.2025.01.007

基于深度强化学习的量子奥托循环性能优化

李建松 1, 李海 1*, 于文莉 2, 郝亚明 1

1 山东工商学院信息与电子工程学院, 山东烟台 264005; 2 山东工商学院计算机科学与技术学院, 山东烟台 264005

收稿日期:2023-02-06 修回日期:2023-04-07 出版日期:2025-01-28 发布日期:2025-01-28
通讯作者: E-mail: shenghuo2003@126.com E-mail:E-mail: shenghuo2003@126.com
作者简介:李建松 ( 1996 - ), 江苏淮安人, 研究生, 主要从事机器学习与量子热力学方面的研究。E-mail: ljs1019@foxmail.com
基金资助:
国家自然科学基金项目 (11547036), 山东省自然科学基金青年项目 (ZR2011FL009), 烟台市科技创新发展计划基金项目 (2022JCYJ044)

Performance optimization of quantum Otto cycle via deep reinforcement learning

LI Jiansong 1 , LI Hai 1*, YU Wenli 2 , HAO Yaming 1

1 School of Information and Electronic Engineering, Shandong Technology and Business University, Yantai 264005, China; 2 School of Computer Science and Technology, Shandong Technology and Business University, Yantai 264005, China

Received:2023-02-06 Revised:2023-04-07 Published:2025-01-28 Online:2025-01-28

摘要/Abstract

摘要： 针对通常情况下实现高性能的绝热捷径量子奥托循环 (QOC) 需要施加复杂调控场的难题, 研究了实验上相对便于操控的线性驱动场下QOC的性能特征。利用基于策略函数的深度强化学习, 对以单量子比特为工质的QOC 膨胀与压缩过程的附加驱动场进行优化, 实现了线性驱动场下高性能的QOC。与非绝热自由演化方案下的QOC对比, 优化附加驱动方案下的QOC在输出功、功率以及效率方面都表现出显著的优越性。特别是在较短循环周期中, 自由演化方案下的QOC因大量不可逆功的产生, 完全抑制了正功的输出, 然而优化驱动方案下的QOC却仍能正常运行(有正功输出)。本工作初步检验了深度强化学习在优化量子热机性能中的有效性。

关键词: 量子热力学, 量子奥托循环, 深度强化学习, 附加驱动场, 功率与效率

Abstract: In response to the challenge that complicated control fields are generally required for realizing the high-performance shortcuts to adiabaticity quantum Otto cycle (QOC), the performance characteristics of QOC under linear driving field which is easy to manipulate in experiment, are studied in this work. Using the strategy-based deep reinforcement learning, the driving field added during the expansion and compression processes of QOC with single qubit as the working medium is optimized, and then the high-performance QOC under linear driving field can be realized. Compared with the scheme of QOC with the non-adiabatic free evolution, the QOC under the optimized additional driving scheme exhibits significant advantages in the output work, power and efficiency. Especially, in the case of shortcycle period, for the QOC under free evolution scheme, the output of positive work is completely suppressed due to the generation of a large amount of irreversible work, while the QOC under the optimized driving scheme can still operate normally (with output positive work). This work preliminarily tests the validity of deep reinforcement learning in optimizing the performance of quantum engine.

Key words: quantum thermodynamics, quantum Otto cycle, deep reinforcement learning, additional driving field, power and efficiency

中图分类号:

O413.1

李建松, 李海, 于文莉, 郝亚明 . 基于深度强化学习的量子奥托循环性能优化[J]. 量子电子学报, 2025, 42(1): 70-0.

LI Jiansong , LI Hai , YU Wenli , HAO Yaming . Performance optimization of quantum Otto cycle via deep reinforcement learning[J]. Chinese Journal of Quantum Electronics, 2025, 42(1): 70-0.

参考文献

[1] Chen X, Lizuain I, Ruschhaupt A, et al. Shortcut to adiabatic passage in two-and three-level atoms[J]. Physical review letters, 2010, 105(12): 123003.

[2] Guéry-Odelin D, Ruschhaupt A, Kiely A, et al. Shortcuts to adiabaticity: Concepts, methods, and applications[J]. Reviews of Modern Physics, 2019, 91(4): 045001.

[3] Hegade N N, Paul K, Ding Y, et al. Shortcuts to Adiabaticity in Digitized Adiabatic Quantum Computing[J]. Physical Review Applied, 2021, 15(2): 024038.

[4] Jarzynski C, Deffner S, Patra A, et al. Fast forward to the classical adiabatic invariant[J]. Physical Review E, 2017, 95(3): 032122.

[5] Patra A, Jarzynski C. Semiclassical fast-forward shortcuts to adiabaticity[J]. Physical Review Research, 2021, 3(1): 013087.

[6] Abah O, Paternostro M, Lutz E. Shortcut-to-adiabaticity quantum Otto refrigerator[J]. Physical Review Research, 2020, 2(2): 023120.

[7] Dupays L, Egusquiza I, Del Campo A, et al. Superadiabatic thermalization of a quantum oscillator by engineered dephasing[J]. Physical Review Research, 2020, 2(3): 033178.

[8] Funo K, Lambert N, Karimi B, et al. Speeding up a quantum refrigerator via counterdiabatic driving[J]. Physical Review B, 2019, 100(3): 035407.

[9] Li L, Li H, Yu W, et al. Shortcut-to-adiabaticity quantum tripartite Otto cycle[J]. Journal of Physics B: Atomic, Molecular and Optical Physics, 2021, 54(21): 215501.

[10] Hartmann A, Mukherjee V, Niedenzu W, et al. Many-body quantum heat engines with shortcuts to adiabaticity[J]. Physical Review Research, 2020, 2(2): 023145.

[11] Zhang X-M, Wei Z, Asad R, et al. When does reinforcement learning stand out in quantum control? A comparative study on state preparation[J]. npj Quantum Information, 2019, 5(1): 1-7.

[12] He R-H, Wang R, Nie S-S, et al. Deep reinforcement learning for universal quantum state preparation via dynamic pulse control[J]. EPJ Quantum Technology, 2021, 8(1): 29.

[13] Sgroi P, Palma G M, Paternostro M. Reinforcement learning approach to nonequilibrium quantum thermodynamics[J]. Physical Review Letters, 2021, 126(2): 020601.

[14] Khait I, Carrasquilla J, Segal D. Optimal control of quantum thermal machines using machine learning[J]. Physical Review Research, 2022, 4(1): L012029.

[15] Erdman P A, Rolandi A, Abiuso P, et al. Pareto-optimal cycles for power, efficiency and fluctuations of quantum heat engines using reinforcement learning[J]. arXiv preprint arXiv:2207.13104, 2022.

[16] Plastina F, Alecce A, Apollaro T J, et al. Irreversible work and inner friction in quantum thermodynamic processes[J]. Physical review letters, 2014, 113(26): 260601.

[17] Sutton R S, Barto A G. Reinforcement Learning: An Introduction[J]. IEEE Transactions on Neural Networks, 1998, 9(5): 1054-1054.

基于深度强化学习的量子奥托循环性能优化

Performance optimization of quantum Otto cycle via deep reinforcement learning

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价

[1]	刘学铭 , 陈永聪 , 敖平. 热噪声环境下的量子调控优化[J]. 量子电子学报, 2024, 41(1): 103-112.
[2]	王升斌, 窦猛汉, 吴玉椿, 郭国平, 郭光灿 . 分布式量子计算研究进展[J]. 量子电子学报, 2024, 41(1): 1-25.
[3]	江俊勤∗, 涂菁. 多势阱中电子的束缚态与能带形成研究[J]. 量子电子学报, 2021, 38(3): 316-326.
[4]	杨春蕾普杰信董永生刘中华梁灵飞. 一种基于量子机制的图像显著性检测方案[J]. J4, 2017, 34(3): 305-315.
[5]	王晓秦猛. 内禀退相干条件下非线性相互作用对量子纠缠和保真度的影响[J]. J4, 2016, 33(5): 561-565.
[6]	许精明阮越. 3-puzzle量子计算的酉变换矩阵及逻辑线路[J]. J4, 2016, 33(4): 469-475.
[7]	许精明阮越. 基于量子衍生算法的8-puzzle问题分析[J]. J4, 2015, 32(4): 459-465.
[8]	王乐刘莹王岩岩赵生妹. 基于重叠法的(49,1,9)量子码CNOT门扩展矩形容错构造[J]. J4, 2014, 31(2): 179-185.
[9]	程建民梁修东朱正芳吴卫锋许业军. 反常玻色产生算符的本征右矢[J]. J4, 2014, 31(2): 135-140.
[10]	王岩岩,刘莹，赵生妹. 基于隐形传态的量子稳定子码容错编码门构造方法研究[J]. J4, 2013, 30(6): 752-758.
[11]	孟婧王海龙龚谦封松林. 外加电场对GaN/AlxGa1-xN双量子阱中性施主束缚能的影响[J]. J4, 2013, 30(3): 360-366.
[12]	史强闫珂柱王晓丽刘志泉. 三结磁通量子比特中处于亚稳态粒子的热逃逸和量子隧穿[J]. J4, 2012, 29(6): 723-728.
[13]	秦猛，李延标，白忠. 不同方向非均匀磁场对Heisenberg XYZ 链中热纠缠的影响[J]. J4, 2012, 29(4): 434-440.
[14]	颜伟倪林. 基于量子联合测量的紧框架构造方法研究[J]. J4, 2011, 28(1): 44-51.
[15]	张仲周波王培吉陶冶薇. 各向异性n维耦合谐振子能量本征值的代数解法[J]. J4, 2009, 26(4): 405-412.