Health State Prognostics Study Based on Extreme Learning Machine

As smart manufacturing's booming, prognostics and health management of equipment becomes more and more important which has emerged as an intelligent support to improve the availability of manufacturing system. Industry equipment consists of many mechanical components, and rolling bearing is one of the most generally applied vital parts in rotating machinery. The smooth operation of the whole system relies on the health of every single part in the system like rolling bearings. Therefore, it is crucial to monitor the health status of key mechanical components of a system. In this paper, a health study of rolling bearing is proposed where ensemble empirical mode decomposition (EEMD) is selected as the signal process method for feature extraction and extreme learning machine (ELM) neural network is implemented for health status prediction. To validate the proposed approach corresponding experimental study is carried out and the results prove the feasibility.