Graduate School of System Informatics, Department of Computational Sciences, Kobe University, Nada-ku, Kobe, Japan.
J Chem Phys. 2013 Apr 28;138(16):164126. doi: 10.1063/1.4802766.
We propose a novel quantum Monte Carlo method in configuration space, which stochastically samples the contribution from a large secondary space to the effective Hamiltonian in the energy dependent partitioning of Löwdin. The method treats quasi-degenerate electronic states on a target energy with bond dissociations and electronic excitations avoiding significant amount of the negative sign problem. The performance is tested with small model systems of H4 and N2 at various configurations with quasi-degeneracy.
我们提出了一种新的组态空间量子蒙特卡罗方法，该方法在洛文的能量相关分区中，通过随机抽样从大的二次空间中获取对有效哈密顿量的贡献。该方法处理具有键离解和电子激发的准简并电子态，避免了大量的负号问题。在具有准简并性的各种构型下，我们使用 H4 和 N2 的小模型系统对该方法的性能进行了测试。
