Modified shuffled frog-leaping algorithm enhanced self-adaption band-pass filtering method for fault diagnosis of screw compressor

Abstract

The invention discloses a modified shuffled frog-leaping algorithm enhanced self-adaption band-pass filtering method for fault diagnosis of a screw compressor. The method includes: 1) performing EEMDprocessing on a collected vibrating signal, calculating the relative kurtosis value of each IMF component, picking the maximum and sub-maximum components of the relative kurtosis value and performingsignal reconstruction; 2) performing modified SFLA-based self-adaption band-pass filtering processing on the reconstructed signal, and precisely cutting-out a high-frequency band signal being rich infault information; 3) performing Hilbert envelope demodulation analysis on the filtered signal, performing spectral analysis on the demodulated signal, and finally diagnosing the fault of the screw compressor. In the invention, firstly, by means of the relative kurtosis value, IMF component reconstruction is carried out to obtain new signals, so that the fault information is maintained as most aspossible and influence on feature extraction due to noise and false component is avoided; secondly, by means of a self-adaption band-pass filter enhanced by the modified SFLA, the reconstructed signalis subjected to the self-adaption band-pass filtering, so that central frequency and bandwidth of band-pass filtering are optimized, and precision of the fault diagnosis is increased.

Description

technical field [0001] The invention relates to the field of fault diagnosis of mechanical equipment, in particular to an improved shuffling leapfrog algorithm enhanced self-adaptive band-pass filter method for fault diagnosis of screw compressors. Background technique [0002] Screw compressor is a double-shaft rotary compressor that works according to the principle of volume change. Compared with piston compressors and centrifugal compressors, screw compressors are simple in structure, reliable in operation and good in volume efficiency. With a series of unique advantages, it is widely used in aerodynamics, refrigeration and air conditioning, and various petrochemical processes. As the core part of a large-scale pressure system, if the operating status of the compressor cannot be judged timely and accurately, the probability of sudden failure will increase, which will affect the normal operation and service life of the unit, and even cause greater economic losses. loss. ...

AI Technical Summary

Problems solved by technology

[0003] However, the gap between the parts of the compressor is extremely small and the movement is fast, and the damage of the parts is judged only by the display of some instruments and actual experience, and its accuracy and effectiveness are extremely poor

In addition, the early fault characteristics of the compressor are very weak. Under actual working conditions, periodic pulse excitations such as damage faults and the eccentricity of the original mass of the rotor will excite the resonance frequencies of each order of the compressor system, and the high frequency band of the signal is mixed with a lot of noise. Impulse components that cause fault information to be easily overwhelmed by noise

For compressor vibration signals that are severely interfered by natural periodic pulses, if only a single signal processing method is used, the demodulation effect is poor, the noise reduction effect is not good, and the fault feature information extraction effect is often not ideal.

Images