A new collaborative fault identification strategy using multivariate hierarchical dispersion entropy

Abstract

This article presents a fault recognition strategy using multivariate hierarchical dispersion entropy to monitor the conditions of rolling bearing. First, the vibration data would be measured from multi-channel sensors synchronously. Then, the proposed mvHDE is employed to capture fault information from the collected data. Finally, the fault features are input into the ELM classifier to automatically identify fault types of bearing. The feasibility and effectiveness of the presented intelligent fault diagnosis schemes are verified through experimental studies.