Stick-slip vibration in water-lubricated bearing-shaft system simulated in persistent homology-based machine learning model

Stick-slip vibration images of water-lubricated rubber stern bearing are collected by using machine vision technology. Then these images are analyzed by the methods of persistent homology-based machine learning (PHML). During this analysis, the corresponding barcode is obtained by calculating the homology of the simplicial complex of the vibration images, and the topological characteristics of the vibration images are obtained based on the barcode images, then the support vector machine (SVM) learning is used to study the topological features, and finally the classification and identification of the stick-slip vibration of water-lubricated rubber stern tube bearing are completed. The results have shown that the length of the longest 1D Betti number is closely related to vibration value. Based on these data, it is possible to use the warning beep effectively, create an intelligent description of the beep process, and provide a new idea for simulating stick-slip vibration in the stern bearing.

Automated summary

The study utilized machine vision technology and persistent homology-based machine learning methods to analyze the stick-slip vibration of water-lubricated rubber stern bearings. By utilizing support vector machine learning, the analysis led to successful classification and identification of the vibration images, with the discovery of a close relationship between vibration value and the length of the 1D Betti number.