Prediction of Remaining Useful Life of Aero-engine Based on Network and Similarity

Youpeng Wan, Wenjin Zhu, Shubin Si

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Aeroengine is a complex system consisting of various components, which has high economic benefits and research value. Accurately evaluating the performance status of aero-engines has become a hot issue in current aero-engine research, and it plays an important role in the maintenance and storage of aero-engines. In this paper, a network model is constructed based on the aero-engine sensors data and the correlation coefficient of sensors. A method for predicting the remaining useful life (RUL) of aero-engines based on the change data of average network node strength and similarity is proposed. Through the node strength to analyze the change of the network node correlation, and the change law of the sensors network overall correlation. The RUL of aero-engines can be predicted accurately. It is found that the correlation between sensors generally increases uniformly at the end of the engine life.

Original languageEnglish
Title of host publication2022 Global Reliability and Prognostics and Health Management Conference, PHM-Yantai 2022
EditorsWei Guo, Steven Li
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781665496315
DOIs
StatePublished - 2022
Event2022 Global Reliability and Prognostics and Health Management Conference, PHM-Yantai 2022 - Yantai, China
Duration: 13 Oct 202216 Oct 2022

Publication series

Name2022 Global Reliability and Prognostics and Health Management Conference, PHM-Yantai 2022

Conference

Conference2022 Global Reliability and Prognostics and Health Management Conference, PHM-Yantai 2022
Country/TerritoryChina
CityYantai
Period13/10/2216/10/22

Keywords

  • aero-engines
  • networks
  • prediction
  • remaining useful life (RUL)

Fingerprint

Dive into the research topics of 'Prediction of Remaining Useful Life of Aero-engine Based on Network and Similarity'. Together they form a unique fingerprint.

Cite this