Alzate-Cardona J D, Sabogal-Suárez D, Evans R F L, Restrepo-Parra E
Departamento de Física y Química, Universidad Nacional de Colombia, Sede Manizales, A.A. 127, Manizales, Colombia.
J Phys Condens Matter. 2019 Mar 6;31(9):095802. doi: 10.1088/1361-648X/aaf852. Epub 2018 Dec 12.
We present an adaptive algorithm for the optimal phase space sampling in Monte Carlo simulations of 3D Heisenberg spin systems. Based on a golden rule of the Metropolis algorithm which states that an acceptance rate of [Formula: see text] is ideal to efficiently sample the phase space, the algorithm adaptively modifies a cone-based spin update method keeping the acceptance rate close to [Formula: see text]. We have assessed the efficiency of the adaptive algorithm through four different tests and contrasted its performance with that of other common spin update methods. In systems at low and high temperatures and anisotropies, the adaptive algorithm proved to be the most efficient for magnetization reversal and for the convergence to equilibrium of the thermal averages and the coercivity in hysteresis calculations. Thus, the adaptive algorithm can be used to significantly reduce the computational cost in Monte Carlo simulations of 3D Heisenberg spin systems.
我们提出了一种用于三维海森堡自旋系统蒙特卡罗模拟中最优相空间采样的自适应算法。基于 metropolis 算法的一条黄金法则，即接受率为[公式：见原文]对于有效采样相空间是理想的，该算法自适应地修改基于锥的自旋更新方法，使接受率接近[公式：见原文]。我们通过四个不同的测试评估了自适应算法的效率，并将其性能与其他常见的自旋更新方法进行了对比。在低温、高温和各向异性系统中，自适应算法在磁化反转以及热平均值收敛到平衡和磁滞计算中的矫顽力方面被证明是最有效的。因此，自适应算法可用于显著降低三维海森堡自旋系统蒙特卡罗模拟中的计算成本。
