Rotational Stall Feature Recognition Of Axial Flow Compressor Based On Sound Pressure Signal

As a kind of unsteady flow phenomenon in compressor,rotating stall is one of the hottest and most difficult points in fluid machinery research field.This paper takes a four-stage axial flow compressor as the research object and carries out in-depth research from the aspects of data processing method and on-line monitoring.In this paper,a series of data processing method is carried out on the collected sound pressure signal to extract the stall characteristics in the data,and the rotation stall characteristic frequency in the signal spectrum diagram is identified through statistical analysis.The main research content is as follows:(1)This paper first introduces the working principle of axial flow compressor and the research process of rotating stall,and summarizes the development of EMD algorithm and wavelet transform.(2)Based on the MPE-CEEMD algorithm,this method is applied to the feature extraction of rotating stall of axial flow compressor to eliminate the abnormal components such as intermittent signal,pulse signal and interference signal to the maximum extent.(3)In order to shorten the running time of the algorithm,a novel algorithm for feature extraction of rotation stall of wavelet soft threshold denoising combined with optimal wavelet tree is proposed.In the time domain,the algorithm well reflects the characteristics of the compressor rotating stall,and does not lose any key information in the frequency domain,and the running time of the algorithm is greatly shortened,which is conducive to the on-line monitoring of rotating stall state.(4)The performance of MPE-CEEMD and wavelet soft threshold combined with optimal wavelet tree denoising is compared.The results show that the running time of the joint optimal wavelet tree algorithm based on the soft threshold of wavelet is much shorter than that of MPE-CEEMD algorithm,but the other index values are worse than that of MPE-CEEMD algorithm.(5)A data acquisition and rotary stall recognition system based on sound pressure signal is developed.Advances have been made in the feature extraction of rotating stall of axial flow compressor based on the sound pressure signal at the outlet housing.