Exploring mono-fractal characteristics of dynamic pressure at exit of centrifugal compressor based on EWT

基于EWT的离心压缩机出口动态压力单重分形特征研究

¹ School of Mechanical Engineering, Northwestern Polytechnical University, Xi’an 710072, China
² State Key Laboratory of Compressor Technology, Hefei General Machinery Research Institute Co., Ltd, Hefei 230031, China

Received: 18 October 2021

Abstract

This paper studies the dynamic pressure at the outlet of the 800 kW centrifugal compressor. It uses the single-fractal method and the empirical wavelet transform to analyze the nonlinear characteristics of the centrifugal compressor under different working conditions. First, it analyzes the characteristics of the dynamic pressure waveform when the centrifugal compressor enters the surge state from the steady state through the transition process. It also uses the empirical wavelet transform to extract the modal components of the pressure waveform, which is reconstructed according to the correlation coefficient. Second, it studies the characteristics of the autocorrelation function of the reconstructed signal under different numbers of hysteresis points k. Finally, it studies the Hurst parameters of the original signal and the reconstructed signal to identify the working state of the centrifugal compressor. The results show that the autocorrelation function of the reconstructed signal can effectively reflect the working state of the centrifugal compressor. Compared with the original signal, it is easier for the Hurst parameter of the reconstructed signal to identify the surge state of the centrifugal compressor.

摘要

以800 kW离心压缩机的出口动态压力为研究对象, 采用经验小波变换并结合单重分形, 分析系统在不同工况下的非线性特征。分析系统从稳定状态经过过渡过程进入喘振状态的出口动态压力波形特征, 采用经验小波变换提取波形的模态分量并根据相关系数进行波形重构; 研究重构信号在不同迟滞点数k下的自相关函数特征; 对比分析原始信号及重构信号的Hurst参数, 以识别系统的工作状态。研究结果显示: 重构信号的自相关函数可以有效地反映系统的工作状态。此外, 相对于原始信号, 重构信号的Hurst参数更易识别出系统的喘振状态。