Study of Fault Diagnosis for Rolling Bearing Based on Clustering Algorithms

Rolling bearing is a common key component in mechanical equipment. Fault diagnosis for rolling bearing is an important guarantee for safe and reliable operation of mechanical equipment. This paper aims to use pattern recognition to analyze the vibration data of the bearing for fault diagnosis. Time domain features of the diagnostic signal were firstly extracted, then the K-means clustering algorithm and K-Nearest Neighbor (KNN) algorithm were used in combination for fault diagnosis, and the advantages and disadvantages of the two algorithms were comparatively analyzed. They show good classification accuracy in fault diagnosis of rolling bearings and have a good research and application prospect.