Original scientific paper

https://doi.org/10.21278/TOF.481054223

Composite Fault Diagnosis in Rotating Machinery Based on Multi-Feature Fusion

Nai-quan Su orcid.org/0000-0002-4220-0540 ; Guangdong Provincial Key Lab of Petrochemical Equipment and Fault Diagnosis, Guangdong University of Petrochemical Technology, Maoming, ChinaHigh-Tech Institute of Xi’an, Xi’an, China
Qing-hua Zhang ; Guangdong Provincial Key Lab of Petrochemical Equipment and Fault Diagnosis, Guangdong University of Petrochemical Technology, Maoming, China
Yi-dian Chen ; Guangdong Provincial Key Lab of Petrochemical Equipment and Fault Diagnosis, Guangdong University of Petrochemical Technology, Maoming, China
Xiao-xiao Chang ; Guangdong Provincial Key Lab of Petrochemical Equipment and Fault Diagnosis, Guangdong University of Petrochemical Technology, Maoming, China
Yang Liu ; Guangdong Provincial Key Lab of Petrochemical Equipment and Fault Diagnosis, Guangdong University of Petrochemical Technology, Maoming, China

Abstract

The rotating machinery working in complex environments of petrochemical units often develops composite faults and its vibration signal exhibits multicoupling, fuzziness, and nonlinearity, making it difficult to effectively diagnose composite faults. This paper proposes a composite fault diagnosis for rotating machinery based on multi-feature fusion. This method firstly extracts the time domain, the frequency domain and the dimensionless feature information, using the correlation analysis and normalization to obtain bodies of evidence with different features. Then, according to the fusion rules of the evidence theory, the synthesis of different bodies of evidence is completed. Finally, the feasibility of the proposed method is verified. The experimental results show that the accuracy of the proposed method exceeds 90%, thus it has been shown that the composite fault diagnosis of rotating machinery in petrochemical units is effective.

Keywords

composite fault diagnosis; time domain; frequency domain; high-value dimensionless; feature fusion

Hrčak ID:

313729

URI

https://hrcak.srce.hr/313729

Publication date:

1.1.2024.

Visits: 201 *