Motor Current Signal Analysis Based on Machine Learning for Centrifugal Pump Fault Diagnosis

Samir Alabied, Alsadak Daraz, Khalid Rabeyee, Ibrahim Alqatawneh, Fengshou Gu, Andrew D. Ball

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

8 Citations (Scopus)

Abstract

Centrifugal pumps are widely used in various industrial processes, such as petrochemical production and power plants. It is important to develop an accurate and cost-effective diagnosis approach for ensuring effective condition monitoring methods. In this paper, the motor current signal has been investigated for the condition monitoring of the centrifugal pump. In specific, a combined approach is proposed based on intrinsic time scale decomposition (ITD) for feature extraction and Support Vector Machine (SVM) for classifying the health conditions. The classification accuracy of SVM was improved significantly when the ITD method is applied to extract the diagnostic features. The results have shown that ITD is efficient and effective for extracting the most informative features from the motor current signals. Also, this indicates that the proposed diagnostic method based on ITD with SVM is more effective to classify the simulated faults of the centrifugal pump.

Conference

Conference25th IEEE International Conference on Automation and Computing
Abbreviated titleICAC 2019
Country/TerritoryUnited Kingdom
CityLancaster
Period5/09/197/09/19
Internet address

Fingerprint

Dive into the research topics of 'Motor Current Signal Analysis Based on Machine Learning for Centrifugal Pump Fault Diagnosis'. Together they form a unique fingerprint.

Cite this