arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.04253quant-phphysics.chem-ph
收藏

构建基于参数标准的紧凑ADAPT单元酉耦合簇算子

Constructing Compact ADAPT Unitary Coupled-Cluster Ansatz with Parameter-Based Criterion

Institute of Software, Chinese Academy of Sciences(中国科学院软件研究所);MindSpore Quantum Special Interest Group(MindSpore量子特别兴趣组);College of Physical Science and Technology, Xinjiang University(新疆大学物理科学与技术学院);Arclight Quantum Computing Inc.(阿克莱特量子计算有限公司)

Runhong He, Xin Hong, Qiaozhen Chai, Ji Guan, Junyuan Zhou, Arapat Ablimit, Guolong Cui, Shenggang Ying

AI总结:

本文提出Param-ADAPT-VQE算法,通过参数标准选择激发算子,减少冗余和测量成本,提升计算精度和可扩展性。

摘要:

适应性导数组装伪 Trotter 变分量子本征求解器(ADAPT-VQE)是一种有前途的混合量子-经典算法,用于分子基态能量计算,但其实际可扩展性受到冗余激发算子和过高的测量成本的阻碍。为解决这些挑战,我们提出了Param-ADAPT-VQE,一种新的改进算法,该算法基于参数标准选择激发算子,而不是传统的梯度标准。这种策略有效避免了冗余算子。我们进一步开发了子哈密顿量技术,并集成了热启动VQE优化策略,实现了测量成本的显著降低。在典型分子系统上的数值实验表明,Param-ADAPT-VQE在计算精度、算子规模和测量成本方面均优于原始ADAPT-VQE。此外,我们的方案保留了ADAPT-VQE的基本框架,因此与各种修改版本完全兼容,能够在特定方面进一步提高性能。本工作为ADAPT-VQE提供了高效且可扩展的改进,缓解了阻碍其在分子量子化学领域实际应用的核心障碍。

英文摘要:

The adaptive derivative-assembled pseudo-trotter variational quantum eigensolver (ADAPT-VQE) is a promising hybrid quantum-classical algorithm for molecular ground state energy calculation, yet its practical scalability is hampered by redundant excitation operators and excessive measurement costs. To address these challenges, we propose Param-ADAPT-VQE, a novel improved algorithm that selects excitation operators based on a parameter-based criterion instead of the traditional gradient-based metric. This strategy effectively eludes redundant operators. We further develop a sub-Hamiltonian technique and integrate a hot-start VQE optimization strategy, achieving a significant reduction in measurement costs. Numerical experiments on typical molecular systems demonstrate that Param-ADAPT-VQE outperforms the original ADAPT-VQE in computational accuracy, ansatz size, and measurement costs. Furthermore, our scheme retains the fundamental framework of ADAPT-VQE and is thus fully compatible with its various modified versions, enabling further performance improvements in specific aspects. This work presents an efficient and scalable enhancement to ADAPT-VQE, mitigating the core obstacles that impede its practical implementation in the field of molecular quantum chemistry.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑