77 / 2023-09-09 12:07:50
Early Failure Point Monitoring for Remaining Useful Life Estimation of Bearings Using Transfer Learning Enabled T2 Control Chart
Early failure points, Statistical Process Control, Transfer learning, varying working conditions
Final Paper
Xianqi Wu / Jiangsu University
Heng Lu / Jiangsu University
Chenhao Yang / Jiangsu University
薇 樊 / 江苏大学
Chao Chen / Jiangsu University
Due to the unique characteristics of bearings under varying working conditions, the control chart technologies in bearing early failure monitoring (BEFM) are facing challenges. In order to address this issue, a kind of covariance transfer enabled T2 control chart (CTET2CC) is proposed to monitor early failure points with limited historical training-data. Firstly, source domain profile data are monitored using T² control charts to identify controlled statistics. Secondly, the method of covariance transfer is proposed to extract the potential controlled information as shared data across domains. Finally, a T² control chart is constructed to monitor the profile data of target domain based on the transferable  statistics. The results verify that the proposed method is expected to be an effective technique for BEFM.

 
Important Date
  • Conference Date

    Nov 02

    2023

    to

    Nov 04

    2023

  • Dec 15 2023

    Draft paper submission deadline

  • Dec 20 2023

    Registration deadline

Sponsored By
IEEE Instrumentation and Measurement Society
Xidian University