基于投影纠缠对态算法优化的研究

Optimization of the projected entangled pair state algorithm for quantum systems

二维强关联电子量子格点系统的投影纠缠对态(PEPS)算法是数值计算领域中研究二维强关联电子量子格点系统最为重要的张量网络算法.基于PEPS算法研究二维量子XYX模型与二维量子Ising模型,本文对PEPS算法进行了一些优化和改进研究,这些优化和改进主要体现在如何进行PEPS张量的更新与如何进行物理观测量的计算这两个方面,从而可以大大提高计算资源的利用.因而优化和改进后的PEPS算法可为研究热力学极限下的二维强关联电子量子格点系统的量子相变和量子临界现象提供一种更有效的强大的工具.

In the numerical calculation, the projected entangled pair state（PEPS） algorithm is the most important tensor network algorithm for two-dimensional strongly correlated electron quantum lattice system. In this paper, the optimization of PEPS for two-dimensional quantum system is discussed. An optimization connection between how to update the PEPS tensor and how to calculate the physical observable is investigated, for the tensor network algorithm based on the PEPS representation, which can greatly improve the utilization of computing resources. In this case, optimized PEPS algorithm, as a powerful tool, can be used to study quantum phase transitions and quantum critical phenomena in the thermodynamic limit of the two-dimensional strongly correlated electron quantum lattice system. Of course,optimization of PEPS algorithm program has many other applications, such as adding U（1） and SO（2） symmetry in PEPS algorithm, etc.