On-line fault diagnosis of rolling bearing based on machine learning algorithm

作者：Jinmeng Sun, Zhongqing Yu, Haiya Wang · 发表于：2020 5th International Conference on Information Science, Computer Technology and Transportation (ISCTT) · 年份：2020 · DOI：10.1109/isctt51595.2020.00075 · 被引用次数：8 · 研究领域：Engineering Diagnostics and Reliability、Machine Fault Diagnosis Techniques、Gear and Bearing Dynamics Analysis

In order to realize the predictive maintenance of rolling bearings in industry, this paper proposes an online fault diagnosis method for rolling bearings based on three machine learning algorithms. The method mainly includes two steps: establishing a fault diagnosis model and online fault diagnosis. Firstly, preprocess the collected bearing vibration data, and then train and optimize the fault diagnosis model, and finally realize online fault diagnosis. The experimental results show that, compared with the traditional bearing fault diagnosis method, the online fault diagnosis of the bearing by the machine learning method is simpler and has a better diagnosis effect.