Multiscale entropy analysis of the velocity signal in Impinging Stream Mixer

Abstract

In laboratory the velocity signal of Impinging Stream Mixer (ISM) was measured with Laser Doppler Anemometer. The velocity signal contains a lot of complicated information. Multiscale entropy (MSE) algorithm, which provides a way to measure complexity over a range of scales, was first used to analysis the complexity of the velocity time series in ISM. According to the MSE theory, draw up Matlab procedures for scientific computing. It is obtained that the choice of parameter has a strong influence on MSE by comparing MSE under different embedding dimension m and tolerance r. It is shown that the axial flow is in state of being dominant by computing MSE under different sections. To distinguish the velocity signal, MSE under different rotating speed were computed. The results indicate that MSE increases with the augment of the rotating speed, and the complexity increases homogeneously.