Generative Adversarial Network-based Missing Data Handling and Remaining Useful Life Estimation for Smart Train Control and Monitoring Systems

Table 3

Specification of the proposed predictive maintenance framework.

Content	Classification	Issues
Data specification	(i) Data source: TCMS data (2018.6∼2019.05); (ii) Data from the seventh line in subway system, Republic of Korea	Big data
TCMS data specification	(i) Number of attributes: 2643 per one record; Existence of a number of missing values in one record; (ii) Data format: encrypted data	(i) Data decryption is needed; (ii) Missing value handling is needed
Fault/alarm data	(i) Number of attributes: 56; (ii) Data format: encrypted text data	(i) Data decryption is needed
Predictive maintenance framework	(i) Data input: TCMS data; (ii) Output: the estimated RUL; (iii) Mechanism: GAN-based deep neural network	—