Fault feature extraction method and fault feature recognition method of vibration signal of rolling mill chatter mark

Abstract

The invention provides a fault feature extraction method and a fault feature recognition method of a vibration signal of a rolling mill chatter mark, and belongs to the field of rolling mill fault monitoring. The fault feature extraction method analyzes the vibration signal of the chatter mark via autocorrelation demodulation based on a second-order cycle, so unstable vibration signal of chatter mark is processed through frequency demodulation, a demodulated second-order cycle autocorrelation function is sliced according to time domain, and modulation information of the chatter mark vibration signal is kept completely. Therefore, accuracy of fault feature extraction of the vibration signal of the chatter mark is improved. The fault feature recognition method recognizes whether the rolling mill faults exist or not by the form of power spectrum information entropy, influence of rolling mill speed fluctuation needs not to be considered, and the recognition method is simple and accurate.

Description

technical field [0001] The present invention relates to a fault feature extraction method and identification method of rolling mill vibration trace vibration signals, in particular to a fault feature extraction method of rolling mill vibration trace vibration signals based on second-order cyclic autocorrelation function and a method based on power spectrum information entropy Fault identification method of rolling mill chatter vibration signal. Background technique [0002] On-line control of rolling mill vibration and chatter vibration has always been a worldwide technical problem. Most steel companies in the world have more or less rolling mill vibration problems. The vibration problem of the rolling mill not only seriously affects the mechanical properties and precision of the rolling mill, but also the strong vibration of the rolling mill will seriously affect the surface quality of the product and bring huge economic losses to the enterprise. In order to study the caus...

AI Technical Summary

Problems solved by technology

However, there are unavoidable speed fluctuations in the rolling mill, which leads to the measured vibration signal being a non-stationary random signal, and it is also likely to bring about frequency modulation, thus covering up the signal information of the vibration marks of the rolling mill

Images