FWDNet: A Novel Recognition Network for Ferrography Wear Debris Image Analysis | Zendy

Fengguang Jia | Zendy; Haijun Wei | Zendy

AI Assistant Blog Pricing

Home ZAIA Blog

Open Access

FWDNet: A Novel Recognition Network for Ferrography Wear Debris Image Analysis

Author(s) -

Fengguang Jia,

Haijun Wei

Publication year - 2022

Publication title -

wireless communications and mobile computing

Language(s) - English

Resource type - Journals

SCImago Journal Rank - 0.42

H-Index - 64

eISSN - 1530-8677

pISSN - 1530-8669

DOI - 10.1155/2022/6511235

Subject(s) - debris , oil analysis , computer science , condition monitoring , identification (biology) , tribology , pattern recognition (psychology) , artificial intelligence , geology , engineering , petroleum engineering , mechanical engineering , oceanography , botany , electrical engineering , biology

Ferrography wear debris in lubricating oil contains abundant worthy information about the state of the machinery and equipment. In order to develop an online monitoring system based on condition maintenance and fault diagnosis, wear debris needs to be identified automatically. Through various tribological experiments, a dataset of seven kinds of wear debris was established. In this study, DenseNet121 was used as the base network to construct a DCNN model (FWDNet) using the transfer learning method. FWDNet obtained an accuracy of 90.15% through a 10-fold crossvalidation test. The results indicate that FWDNet and DCNN mode is suitable for the identification of wear debris and can be used in actual condition monitoring systems in the future.

The content you want is available to Zendy users.

Already have an account? Click here to sign in.

Having issues? You can contact us here

Accelerating Research