진학프로석·박사 채용 접수 플랫폼

Molecular simulations are essential for exploring the kinetic properties of physico-chemical systems, yet their practical application is often constrained by the overwhelming computational demands of high-dimensional systems. To make these simulations more tractable, dimensionality reduction techniques are employed to focus on the most relevant degrees of freedom, effectively shrinking the size of the system under study. However, this reduction introduces memory effects, where the dynamics of the system become dependent not only on its current state but also on its past evolution. These effects emerge because the discarded degrees of freedom—considered as part of the environment—continue to influence the system through delayed interactions. Addressing this challenge is critical: reducing the dimensionality of simulations while preserving their accuracy requires a robust framework to capture how the environment’s influence can be represented as colored noise, which accounts for temporal correlations in the system’s behavior.

This project will primarily focus on the numerical analysis and implementation of noise extraction from atomistic systems, aiming to model the environment’s influence as colored noise. The study will involve developing and testing computational methods to characterize and incorporate memory effects into reduced-dimensionality simulations. Several test cases will be examined, including ionic systems, pores in minerals, and crystal interfaces, to assess the robustness and generality of the approach. By analyzing these diverse systems, the project seeks to demonstrate how memory effects and colored noise can be systematically integrated into simulations. The results will provide insights into refining computational strategies for large-scale molecular simulations in materials science and computational physics.

The project will involve substantial numerical development, including algorithm design, implementation, and validation against atomistic simulation data.