비선형 보간 기반의 희소점 표현 웨이블렛

Abstract

In this work, a nonlinear interpolation-based sparse point representation wavelets is presented to effectively represent local nonlinear features of a flow field. The method is an interpolation-based wavelets using locally varying orders of accuracy. We employ the sparse point representation to attain fine compression ratio and to reduce computation workload. The order variation algorithm is applied for the purpose of preserving highly nonlinear features such as shock and vortices. To verify the proposed method, we consider several one-dimensional discontinuity problems, a two-dimensional heat conduction, and a disturbed wave of half sinusoidal shape. From the results, it is shown that the nonlinear interpolation wavelets provides us with high accuracy and fine compression ratio in benign flow including local peculiar features. Thus it may be concluded that the nonlinear interpolation wavelets is highly promising for discontinuous CFD data compressions.