arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2406.12207cond-mat.str-elcond-mat.other
收藏

结合投影纠缠对态方法的格林函数蒙特卡洛方法用于阻挫$J_1$-$J_2$海森堡模型研究

The Green's function Monte Carlo combined with projected entangled pair state approach to the frustrated $J_1$-$J_2$ Heisenberg model

He-Yu Lin, Yibin Guo, Rong-Qiang He, Z. Y. Xie, Zhong-Yi Lu

AI总结:

针对张量网络算法计算成本高的问题,提出格林函数蒙特卡洛与投影纠缠对态结合的混合方法,在阻挫$J_1$-$J_2$海森堡模型中提升了基态能量精度并确定了中间参数区的可能相。

摘要:

张量网络算法是一类用于求解量子多体问题的主流数值方法,它能够精准捕捉量子系统固有的纠缠特性,实现对量子态的精确表征。然而,该方法的计算成本极高,尤其是在计算关联函数等物理可观测量时。为克服这一计算难题并提升效率,本文提出将格林函数蒙特卡洛(Green's function Monte Carlo, GFMC)方法与投影纠缠对态(projected entangled pair state, PEPS)拟设相结合。该方法融合了蒙特卡洛的高效特性与张量网络态的无符号问题特性,被证实可有效解决计算瓶颈。为展示其性能,我们将这一混合方法应用于研究正方格子上的反铁磁$J_1$-$J_2$海森堡模型——该模型因在量子蒙特卡洛模拟中存在符号问题而闻名。研究结果表明,采用初级PEPS作为GFMC的引导波函数时,基态能量的计算精度得到了显著提升。通过计算结构因子和自旋-自旋关联函数,我们进一步表征了相图,发现在$0.52 < J_2/J_1 < 0.58$的中间参数范围内可能存在柱状价键态相。这项综合研究凸显了我们所提组合方法的有效性,证明其能够在保证计算效率的同时,精确模拟阻挫量子自旋系统。

英文摘要:

The tensor network algorithm, a family of prevalent numerical methods for quantum many-body problems, aptly captures the entanglement properties intrinsic to quantum systems, enabling precise representation of quantum states. However, its computational cost is notably high, particularly in calculating physical observables like correlation functions. To surmount the computational challenge and enhance efficiency, we propose integrating the Green's function Monte Carlo (GFMC) method with the projected entangled pair state (PEPS) ansatz. This approach combines the high-efficiency characteristics of Monte Carlo with the sign-free nature of tensor network states and proves effective in addressing the computational bottleneck. To showcase its prowess, we apply this hybrid approach to investigate the antiferromagnetic $J_1$-$J_2$ Heisenberg model on the square lattice, a model notorious for its sign problem in quantum Monte Carlo simulations. Our results reveal a substantial improvement in the accuracy of ground-state energy when utilizing a preliminary PEPS as the guiding wave function for GFMC. By calculating the structure factor and spin-spin correlation functions, we further characterize the phase diagram, identifying a possible columnar valence-bond state phase within the intermediate parameter range of $0.52 < J_2/J_1 < 0.58$. This comprehensive study underscores the efficacy of our combined approach, demonstrating its ability to accurately simulate frustrated quantum spin systems while ensuring computational efficiency.

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。

↑