Multi-source and multi-fault condition monitoring based on parallel factor analysis and sequential probability ratio test

Liu Yang,Hanxin Chen,Yao Ke,Yuzhuo Miao,Menglong Li,Lang Huang

doi:10.1186/s13634-021-00730-w

Liu Yang, Hanxin Chen + Show 4 more

Open Access

https://doi.org/10.1186/s13634-021-00730-w

Copy DOI

Abstract

The monitoring of mechanical equipment systems contains an increasing number of complex content, expanding from traditional time, and frequency information to three-dimensional data of the time, space, and frequency information, and even higher-dimensional data containing subjects, experimental conditions. For high-dimensional data analysis, traditional decomposition methods such as Hilbert transform, fast Fourier transformation, and Gabor transformation not only lose the integrity of the data, but also increase the amount of calculation and introduce a lot of redundant information. The phenomenon of feature coupling, aliasing, and redundancy between the mechanical multi-source data signals will cause the inaccuracy of the evaluation, diagnosis, and prediction of industrial production operation status. The analysis of the three-way tensor composed of channel, frequency, and time is called parallel factor analysis (PARAFAC). The properties between the parallel factor analysis results and the input signals are studied through simulation experiments. Parallel factor analysis is used to decompose the third-order tensor composed of channel-time-frequency after continuous wavelet transformation of vibration signal into channel, time, and frequency characteristics. Multi-scale parallel factor analysis successfully extracted non-linear multi-dimensional dynamic fault characteristics by generating the spatial, spectral, time-domain signal loading value and three-dimensional fault characteristic expression. In order to verify the effectiveness of the space, frequency, and time domain signal loading values of the fault characteristic factors generated by the centrifugal pump system after parallel factor analysis, the characteristic factors obtained after parallel factor analysis are used as the SPRT test sequence for identification and verification. The results indicate that the method proposed in this article improves the measurement accuracy and intelligence of mechanical fault detection.

Highlights

In recent years, equipment fault diagnosis, as a new technology that crosses various disciplines, has been developed rapidly and has produced huge economic benefits [1,2,3,4,5,6]
Wenjian Huang et al [11] extracted the characteristic parameters of the vibration signal through time-domain signal analysis, the PCA was used to reduce the amount of data, and the main component with the largest contribution rate was used as the input signal of sequential probability ratio test (SPRT) to evaluate the proposed algorithm
The third-order tensor constructed by multi-channel vibration signals through continuous wavelet transform is decomposed into the series of different modes of channel/frequency/time by the multiscale parallel factor analysis algorithm

Summary

Introduction

Equipment fault diagnosis, as a new technology that crosses various disciplines, has been developed rapidly and has produced huge economic benefits [1,2,3,4,5,6]. Researchers have proposed various effective diagnostic methods to process the collected raw vibration signals of the centrifugal pump, extract effective information, and improve the accuracy of diagnosis. Wenjian Huang et al [11] extracted the characteristic parameters of the vibration signal through time-domain signal analysis, the PCA was used to reduce the amount of data, and the main component with the largest contribution rate was used as the input signal of SPRT to evaluate the proposed algorithm. The traditional decomposition method of processing highdimensional data will lose the integrity of the data and increase the amount of calculation and introduce redundancy [15,16,17,18] These methods of extracting time-frequency characteristics from single-channel signals, such as Fourier transform, cannot reflect the internal relationship of non-linear changes between multi-source channel characteristic signals, nor can they eliminate information interference

Methods

Results

Conclusion