TY - GEN
T1 - Intelligent Predictive Maintenance Strategy for Hybrid Systems Using Model-Data Fusion Approach
AU - Xiao, Chenyu
AU - Zheng, Pai
N1 - Publisher Copyright:
© 2023 IEEE.
PY - 2023/8
Y1 - 2023/8
N2 - This paper proposes an intelligent predictive maintenance framework subjected to multiple faults in hybrid systems. This framework embeds a model-data fusion approach, where the reinforcement extended Kalman filter (REKF) is for model-based fault estimation and the biogeography based optimization-support vector regression (BBO-SVR) is for data-driven remaining useful life (RUL) prediction. The intelligent predictive maintenance framework for hybrid systems can be constructed as follows. Firstly, the REKF is adopted for fault estimation based on state space equations of the hybrid system to identify the degradation processes of faulty components. Then, an event-driven decision module is designed to sequentially judge the states of RUL prediction procedure and maintenance procedure. After that, the BBO-SVR is developed to predict the RULs of faulty components under different health monitoring stages of the hybrid systems, which are determined by the results of event-driven decision module. Finally, the feasibility of the above proposed intelligent predictive maintenance framework will be validated by a case study.
AB - This paper proposes an intelligent predictive maintenance framework subjected to multiple faults in hybrid systems. This framework embeds a model-data fusion approach, where the reinforcement extended Kalman filter (REKF) is for model-based fault estimation and the biogeography based optimization-support vector regression (BBO-SVR) is for data-driven remaining useful life (RUL) prediction. The intelligent predictive maintenance framework for hybrid systems can be constructed as follows. Firstly, the REKF is adopted for fault estimation based on state space equations of the hybrid system to identify the degradation processes of faulty components. Then, an event-driven decision module is designed to sequentially judge the states of RUL prediction procedure and maintenance procedure. After that, the BBO-SVR is developed to predict the RULs of faulty components under different health monitoring stages of the hybrid systems, which are determined by the results of event-driven decision module. Finally, the feasibility of the above proposed intelligent predictive maintenance framework will be validated by a case study.
UR - http://www.scopus.com/inward/record.url?scp=85174392686&partnerID=8YFLogxK
U2 - 10.1109/CASE56687.2023.10260602
DO - 10.1109/CASE56687.2023.10260602
M3 - Conference article published in proceeding or book
AN - SCOPUS:85174392686
T3 - IEEE International Conference on Automation Science and Engineering
BT - 2023 IEEE 19th International Conference on Automation Science and Engineering, CASE 2023
PB - IEEE Computer Society
T2 - 19th IEEE International Conference on Automation Science and Engineering, CASE 2023
Y2 - 26 August 2023 through 30 August 2023
ER -