反应扩散的伸缩式投影多尺度模拟方法

Telescoping Projective Methods for the Multiscale Simulation of Reaction-Diffusion Processes

目前在复杂能源系统模拟研究中所面临的各类反应扩散问题越来越复杂,其在宏观上已不能有效模拟,在细小尺度上由于内存开销巨大与CPU耗时过长实施模拟极为困难。基于此,本文提出一种反应扩散的伸缩式投影多尺度模拟方法(RDTPM)。该方法只需通过反应扩散的细小仿真器计算少量演化步,用伸缩式投影积分方式对仿真器演化得到的反应物浓度进行合理的外推处理,即能准确快速地获取后续演化步的反应物浓度,从而满足反应扩散的研究所需。分别利用动态学蒙特卡洛模拟(KMC)和格子玻尔兹曼方法(LBM)建立反应扩散的细小仿真器,通过对Schlogl、Selkov反应扩散过程的多尺度模拟结果证明所提方法的准确性和高效性。

At present,various reaction-diffusion problems existed in the simulation research of complex energy systems are becoming increasingly complicated,consequently,they could not be effectively conducted at macro-scale level and they are also very difficult to simulate at a fine level because of overhead memory and too much CPU time.For these reasons telesoping projective methods for the multiscale simulation of reaction-diffusion（RDTPM） is proposed.This methods only need to calculate a small amount of evolution steps by fine scale simulators of reaction-diffusion,and then the obtained reactant concentration is reasonably extrapolated using teleprojective integration,the concentration of follow-up evolution steps could be obtained quickly and precisely.The kinetic Monte Carlo（KMC）and lattice Boltzmann method（LBM） are used to establish fine scale simulators of reaction-diffusion,and the multiscale simulation results of Schlogl,Selkov reaction-diffusion processes show that the proposed methods are accurate and efficient.