TY - GEN
T1 - Data Driven based Modeling and Fault Detection for the MATLAB/Simulink Turbofan Engine
T2 - 2022 IEEE Conference on Control Technology and Applications, CCTA 2022
AU - Peng, Chao Chun
AU - Cheng, Yi Ho
N1 - Publisher Copyright:
© 2022 IEEE.
PY - 2022
Y1 - 2022
N2 - In recent years, digital twin-based techniques have been applied in many different engineering fields, including simulation, model design, performance analysis, system prognosis, and so on. However, for complex systems, the associated physical equations and dynamics are sometimes difficult to be constructed using domain knowledge-based derivations. On the contrary, data-driven-based modeling can skip this challenge providing the associated input/output data are available. As a result, this paper presents a modified data-driven method based on the AutoRegressive with eXogenous input (ARX). The Observer/Kalman filter identification (OKID) with eigensystem realization algorithm (ERA) and fast orthogonal search (FOS) are used for system identification in order to create digital twins that can consider the model stability, which most of the data-driven methods have lack of. The condition of the system is determined based on the difference between the real system and digital twins. Finally, a simulation using the MATLAB/SIMULINK Turbofan Engine is taken as a tested black box. Simulation results prove that the proposed method can identify faulty condition.
AB - In recent years, digital twin-based techniques have been applied in many different engineering fields, including simulation, model design, performance analysis, system prognosis, and so on. However, for complex systems, the associated physical equations and dynamics are sometimes difficult to be constructed using domain knowledge-based derivations. On the contrary, data-driven-based modeling can skip this challenge providing the associated input/output data are available. As a result, this paper presents a modified data-driven method based on the AutoRegressive with eXogenous input (ARX). The Observer/Kalman filter identification (OKID) with eigensystem realization algorithm (ERA) and fast orthogonal search (FOS) are used for system identification in order to create digital twins that can consider the model stability, which most of the data-driven methods have lack of. The condition of the system is determined based on the difference between the real system and digital twins. Finally, a simulation using the MATLAB/SIMULINK Turbofan Engine is taken as a tested black box. Simulation results prove that the proposed method can identify faulty condition.
UR - http://www.scopus.com/inward/record.url?scp=85144598271&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85144598271&partnerID=8YFLogxK
U2 - 10.1109/CCTA49430.2022.9966098
DO - 10.1109/CCTA49430.2022.9966098
M3 - Conference contribution
AN - SCOPUS:85144598271
T3 - 2022 IEEE Conference on Control Technology and Applications, CCTA 2022
SP - 498
EP - 503
BT - 2022 IEEE Conference on Control Technology and Applications, CCTA 2022
PB - Institute of Electrical and Electronics Engineers Inc.
Y2 - 23 August 2022 through 25 August 2022
ER -