多尺度优化中kMC仿真计算效率的优化

Optimization of kMC Simulation Computational Efficiency in Multiscale Optimization

为了降低计算代价,实现多尺度优化,本文从微观模型入手,以燃料电池反应-扩散过程为例,在动力学蒙特卡洛法(kinetics Monte Carlo,简称kMC)仿真中引入非结构化自适应建表方法(in situ Adaptive Tabulation,简称ISAT),组成kMC-ISAT混合方法,优化kMC仿真的计算效率。将该混合方法与传统kMC仿真结果进行对比,结果表明采用混合方法能够有效地减少仿真时间,优化计算效率,同时保证仿真的准确性。

In order to reduce the Computational price and realize the muhiscale optimization, from the standpoint of microcosmic model, this paper takes the reacting-diffusion process in the fuel cell for an example and Situ Adaptive Tabulation （ISAT） is introduced to the kinetics Monte Carlo （kMC） simulation, so a kMC-ISAT hybrid method is established to reduce, the computational time of kMC simulation. Compare the result with the corresponding one of kMC simulation. The result demonstrates the hybrid method can reduce the simulation time of kMC simulation and advance the calculation efficiency. At the same time, the accuracy is high enough.