Centrifugal Pumps Fault Diagnosis Using Multivariate Multiscale Symbolic Dynamic Entropy and Logistic Regression

The fault diagnosis of centrifugal pumps is crucial to ensure its safety operation and reduce the maintenance costs. In this paper, a novel framework is established for fault diagnosis of centrifugal pumps. First, the multivariate multiscale symbolic dynamic entropy (MvMSDE) is proposed to extract the fault features from the measured synchronous multi-channel vibration signals. Then, the fault features are taken as the input of logistic regression (LR) to classify different fault types of a centrifugal pump. The effectiveness of the proposed method is validated using the experimental data. Meanwhile, a comparison is conducted between MSDE and MvMSDE. Results show that the proposed method has better performance to recognize different bearing and impeller faults of centrifugal pumps.