Deep learning based fault diagnostics for manufacturing systems considering data imbalance

Due to the recent 4th industrial revolution, manufacturing systems have become intelligent with the fusion of the industrial internet of things (IIoT) and cloud computing. To improve the reliability and availability of intelligent manufacturing systems, data-driven fault diagnostic methods have been paid much attention due to numerous data obtained from the manufacturing systems. Among data-driven approaches, deep learning based fault diagnostic methods have shown outstanding performance. However, the existing deep learning based approaches are limited in the case of an imbalanced dataset. In real-world manufacturing applications, data imbalance is commonly encountered because the manufacturing systems are mostly operated under healthy conditions. To solve this challenge, in this project, we develop a new deep learning model for fault diagnostics of manufacturing systems considering data imbalance problem. The developed model will be further validated using the dataset from the testbed and real-manufacturing system to realize the fault diagnostics for the intelligent manufacturing systems.

Faculty Supervisor:

Chi-Guhn Lee

Student:

Partner:

Seoul National University

Discipline:

Engineering

Sector:

Education

University:

University of Toronto

Program: