A Novel Statistical Time-Frequency Analysis for Rotating Machine Condition Monitoring

In this paper, a new method for time-varying machine condition monitoring is proposed. By Choi-Williams distribution, the interference terms produced by the bilinear time-frequency transform are reduced and the fault signal is processed by the correlation analysis of the Choi-Williams distribution. For machine fault diagnosis, both the feature extractor and classifier are combined to make a decision. It is particularly suited to those who are not experts in the field. Satisfactory results have been obtained from a real example and the effectiveness of the proposed method is demonstrated.

Download Full-text

application based comparison of statistical versus deep learning approaches to reciprocating compressor valve condition monitoring

Annual Conference of the PHM Society ◽

10.36001/phmconf.2021.v13i1.3081 ◽

2021 ◽

Vol 13 (1) ◽

Author(s):

Jason Kolodziej ◽

Jacob Chesnes

Keyword(s):

Deep Learning ◽

Condition Monitoring ◽

Frequency Analysis ◽

Reciprocating Compressor ◽

Learning Approaches ◽

Early Assessment ◽

Valve Seat ◽

Repeated Impact ◽

Time Frequency Analysis ◽

Time Frequency

This paper presents a vibration-based condition monitoring approach for early assessment of valve wear in an industrial reciprocating compressor. Valve seat wear is a common fault mode that is caused by repeated impact and accelerated by chatter. Seeded faults consistent with valve seat wear are installed on the head-side discharge valves of a Dresser-Rand ESH-1 industrial reciprocating compressor. Due to the cyclostationary nature of these units a time-frequency analysis is employed where targeted crank angle positions can isolate externally mounted, non-invasive, vibration measurements. A region-of-interest (ROI) is then extracted from the time-frequency analysis and used to train a suitably sized convolutional neural network (CNN). The proposed deep learning method is then compared against a similarly trained discriminant classifier using the same ROIs where features are extracted using texture and shape image statistics. Both methods achieve > 90% success with the CNN classification strategy nearing a perfect result.

Download Full-text

Machine Condition Monitoring Using Signal Classification Techniques

Journal of Vibration and Control ◽

10.1177/107754603030683 ◽

2003 ◽

Vol 9 (10) ◽

pp. 1103-1120 ◽

Cited By ~ 7

Author(s):

M. Ch. Pan ◽

P. Sas ◽

H. Van Brussel

Keyword(s):

Fault Diagnosis ◽

Condition Monitoring ◽

Classification Performance ◽

Signal Classification ◽

Machine Condition ◽

Window Length ◽

Time Frequency ◽

Post Process ◽

Machine Condition Monitoring ◽

Vibration Signatures

Two signal classification approaches, based on Wigner-Ville distribution and extended symmetric Itakura distance, are proposed to post-process the time-frequency representations (TFRs) of vibration signatures, with the final aim to arrive at an automated procedure of machine condition monitoring. Three synthetical signals are used to evaluate and compare the classification performance of these techniques. Some related computation issues, such as characters of different TFRs and weighted window length, are discussed. Experimental case studies, joint fault diagnosis, are realized.

Download Full-text

An Object-Oriented Machine Condition Monitoring Scheme for Rotating Machinery

Volume 1D: 16th Biennial Conference on Mechanical Vibration and Noise ◽

10.1115/detc97/vib-4235 ◽

1997 ◽

Author(s):

Bernhard P. Bettig ◽

Ray P. S. Han

Keyword(s):

Condition Monitoring ◽

Numerical Models ◽

Object Oriented ◽

Rotating Machinery ◽

Machine Condition ◽

Rotating Machine ◽

Machine Condition Monitoring ◽

Framework Approach ◽

Bearing Stiffness ◽

User Friendly

Abstract The increasing reliability and accuracy of sensors as well as improvements in data acquisition and display have lead to the predictive maintenance method of scheduling machinery overhauls and replacement. Numerical models of the dynamics of a machine may be used to more accurately predict and schedule maintenance by relating variables describing deterioration mechanisms to machine measurements. For instance, vibration measurements may be used to determine bearing stiffness which is then trended. The present paper describes the concepts of a program to interactively examine and predict the health of a rotating machine. To facilitate a user-friendly implementation of the method, the paper also presents an object-oriented framework approach for introducing GUI-based graphics.

Download Full-text