An evaluation of noise reduction algorithms for particle-based fluid simulations in multi-scale applications

M.J Zimoń,R Prosser,D.R Emerson,M.K Borg,D.J Bray,L Grinberg,J.M Reese

doi:10.1016/j.jcp.2016.08.021

Abstract

Filtering of particle-based simulation data can lead to reduced computational costs and enable more efficient information transfer in multi-scale modelling. This paper compares the effectiveness of various signal processing methods to reduce numerical noise and capture the structures of nano-flow systems. In addition, a novel combination of these algorithms is introduced, showing the potential of hybrid strategies to improve further the de-noising performance for time-dependent measurements. The methods were tested on velocity and density fields, obtained from simulations performed with molecular dynamics and dissipative particle dynamics. Comparisons between the algorithms are given in terms of performance, quality of the results and sensitivity to the choice of input parameters. The results provide useful insights on strategies for the analysis of particle-based data and the reduction of computational costs in obtaining ensemble solutions.

Highlights

Numerical simulation is an essential tool for gaining a better understanding of many physical phenomena that can be difficult to describe with analytical methods or experimental studies
The statistical mechanics of complex systems is often analysed with molecular dynamics (MD) [1], Monte Carlo methods, e.g. direct simulation Monte Carlo (DSMC) [2] or dissipative particle dynamics (DPD) [3]; a comprehensive summary of all the modelling strategies can be found in Karniadakis et al [4]
In order to relate the results to an initial noise level, we introduce the overall dimensionless gain in signal-to-noise ratio: Gain = SNRapprox − SNRnoisy, SNRnoisy where SNRnoisy and SNRapprox are the SNR values of the original corrupted signal and de-noised data, respectively

Summary

Introduction

Numerical simulation is an essential tool for gaining a better understanding of many physical phenomena that can be difficult to describe with analytical methods or experimental studies. The statistical mechanics of complex systems is often analysed with molecular dynamics (MD) [1], Monte Carlo methods, e.g. direct simulation Monte Carlo (DSMC) [2] or dissipative particle dynamics (DPD) [3]; a comprehensive summary of all the modelling strategies can be found in Karniadakis et al [4]. These procedures can be used to resolve accurately the dynamics at atomistic, meso- and micro-scales and are widely used to simulate nano/micro flows confined in channels such as carbon nanotubes [5,6]. There is significant potential to apply multi-scale techniques to sociological problems, such as crowd and traffic flow [12]

Objectives

Results

Conclusion