Condition Monitoring And Fault Diagnosis Of Rolling Bearing Based On Vibration Signal

With the development of large industrial equipment, the rolling bearing which is the mosteasily damaged part in rotating machinery affects the operation of the entire system.Mechanical failure caused by the damage of rolling bearings results huge economic loss.Therefore, there is very important and significant to study how to detect and identify thefault of rolling bearings accurately and efficiently.Method of fault detection andidentification which is based on principal component analysis obtains great progress anddevelopment through the related practical research. However, some problems have to besolved and studied which is due to the time-frequency characteristic of vibration signal andthe complexity and the signal correlation processing. This paper tries to study the method offault detection and identification method accurately and wildly applied.This paper studies the acquisition of the platform from the QPZZ-II fault simulation ofrolling bearing vibration signal and realizes the recognition of fault types. This method firstlysolves the one-dimensional time series which is not convenient for the PCA analysis of theproblem by using the C-C phase space reconstruction; then it sets up the PCA model in thespace and realizes the monitoring of the condition of bearings using the T2statistic and SPEstatistic. In order to further analyses of fault types, this paper puts forward fault diagnosismethod based on the multi PCA model and identifies fault types using PCA similarityprinciple. From the simulation results, this paper puts forward that the rolling bearings’ faultdetection and identification method based on vibration signal can detect the fault informationcontained in the bearing data effectively and accurately.