Fault diagnosis of rolling bearing using symmetrized dot pattern and density-based clustering

• Published in: Measurement : journal of the International Measurement Confederation Vol. 152; p. 107293
• Main Authors: Li, Hai, Wang, Wei, Huang, Pu, Li, Qingzhao
• Format: Journal Article
• Language: English
• Published: London Elsevier Ltd 01.02.2020; Elsevier Science Ltd
• Subjects: Bearings; Clustering; Density; Density-based clustering; Diagnostic systems; Fatigue failure; Fault diagnosis; Genetic algorithms; Noise; Noise reduction; Optimization; Roller bearings; Rolling bearing; Signal processing; Symmetrized dot pattern; Symmetrized dot pattern analysis; Vibration; Fault diagnosis; Symmetrized dot pattern analysis; Density-based clustering; Rolling bearing
• ISSN: 0263-2241; 1873-412X
• DOI: 10.1016/j.measurement.2019.107293

•A novel fault diagnosis method using ASDP-DBSCAN for rolling bearing is proposed.•HFGA is adopted to adaptively select SDP-parameters.•An improved DBSCAN is used to generate fault clustering template.•The effectiveness of the proposed method is fully evaluated by experiments. The rolling bearing usually works under complex working conditions, which makes it more susceptible to mechanical failure. The vibration signals are usually complex, nonlinear and non-stationary. This paper proposed a novel diagnosis method for rolling bearing combined with the adaptive symmetrized dot pattern and density-based spatial clustering of applications with noise (ASDP-DBSCAN). Firstly, the SDP technique is briefly introduced and then the vibration signals are reconstructed by the SDP pattern. Secondly, in order to maximize the difference between SDP patterns, a novel parameter optimization method of SDP pattern is presented combined with Hill function and genetic algorithm (HFGA), which is conducive to improve diagnostic accuracy. Then, an improved DBSCAN is used to generate clustering template so as to reduce the effect of noise on diagnostic accuracy. Furthermore, the similarity analysis between clustering template and unknown SDP pattern is used to fault classification. Finally, the proposed method is applied to fault diagnosis for the rolling bearing. The experimental results validate that the proposed method is more effective than other methods for rolling bearing fault diagnosis.