Reciprocating compressor air valve fault diagnosis method based on time-varying singular spectrum

Abstract

The invention relates to a reciprocating compressor air valve fault diagnosis method based on a time-varying singular spectrum, and the method specifically comprises the steps: collecting an air valvevibration acceleration signal and a pressure signal, carrying out the wavelet noise reduction and decomposition of an initial vibration acceleration signal, and removing a pseudo-component; selectinga proper time delay to construct a window function, and introducing time information into the multi-fractal analysis of the vibration signal, thereby obtaining a time-varying singular spectrum; and enabling the drawn time-varying singular spectrogram of each working process to describe the whole and detail information of the vibration signal more finely, so that parameters with large influence difference for a certain fault type are captured, fault feature vectors are convenient to construct, an SVM classifier is convenient to establish, parameters are optimized and set, and finally, the fault diagnosis of the air valve is successfully realized. The problem of fault diagnosis of the air valve of the reciprocating compressor is solved.

Description

Technical field: [0001] The invention relates to the technical field of fault diagnosis of reciprocating machinery, in particular to a method for fault diagnosis of an air valve of a reciprocating compressor based on a time-varying singular spectrum. Background technique: [0002] In the field of petrochemical industry, reciprocating compressors are often used as a power source for gas boosting and transmission media, and are widely used in the operation of oil and gas refining and long-distance oil and gas pipelines. However, due to the complex structure of the reciprocating compressor, the parts are subject to alternating forces, and are often used to transport unstable gases such as natural gas and hydrogen, and the working environment is harsh. Moreover, the defects or failures of some parts of the compressor may cause problems with the mechanical system. The fault frequency of a certain part of the same natural frequency will aggravate the vibration of the entire mechan...

AI Technical Summary

Problems solved by technology

However, the above method only analyzes the vibration signal as a whole, and the obtained results still cannot describe the detailed characteristics of the vibration signal.

[0004] The time-varying characteristics of the vibration signal of the reciprocating compressor system are complex and changeable, and the sampling accuracy and conditions are limited. When calculating the time-varying singularity index of the actual fractal process, if the instantaneous autocorrelation function is directly used for calculation, the calculation amount is huge and it is not easy to be accurate. estimate

Images