Rolling bearing fault diagnosis based on RQA with STD and WOA-SVM

• DOI: 10.1016/j.heliyon.2024.e26141
• Publication Date: 2024-02-09T08:45:42Z
• Authors: Wentao Qiu; Bing Wang; Xiong Hu
• Keywords: H1-99; 0209 industrial biotechnology; Support vector machine; Science (General); Rolling bearing; 02 engineering and technology; Social sciences (General); Q1-390; Standard deviation; Whale optimization algorithm; Recursive quantitative analysis; Fault diagnosis; Research Article

A rolling bearing fault diagnosis method based on Recursive Quantitative Analysis (RQA) combined with time domain feature extraction and Whale Optimization Algorithm Support Vector Machine (WOA-SVM) is proposed. Firstly, the recurrence graph of vibration signal drawn, nonlinear parameters in Standard Deviation (STD) are extracted by recursive quantitative analysis to generate vectors; after that, order construct optimal support vector machine model, used optimize c g parameters. Finally, both standard deviation WOA-SVM model perform bearings. The datasets from Case Western Reserve University Jiangnan were for example analysis, identification accuracy reached 100% 95.00%, respectively. Compared other methods, proposed this paper has higher diagnostic wide practical applicability, risk accidents can be reduced through accurate diagnosis, which also important safety environmental policies. This research originated field mechanical solve problem bearings industrial production, it builds previous explores new methods techniques fill some gaps diagnosis.