TY - JOUR
T1 - Evaluation of Aero-engine Fault Diagnosability Under Multi-source Uncertainty
AU - Sun, Rui Qian
AU - Zhao, Jian Ping
AU - Wang, Yuan Fan
AU - Gou, Lin Feng
N1 - Publisher Copyright:
© The Author(s), under exclusive licence to The Korean Society for Aeronautical & Space Sciences 2025.
PY - 2025
Y1 - 2025
N2 - Aero-engine is faced with various uncertainties while operating. To reduce the conservativeness of diagnosis system in complex environment, it is of great significance to consider the impact of multi-source uncertainty of fault diagnosis systems during the design process, the most effective way of which is to quantify the fault diagnosability. In this paper, the multi-source uncertainty environment, which is composed of aero-engine epistemic uncertainty caused by modeling and stochastic uncertainties of actuator output and sensor measurement in control system, is considered, and the diagnostic evaluation of actuator, sensor and gas-path component faults is carried out, respectively. Based on ITO integral and polynomial chaos expansion, the agent model of output response was established, and the uncertainty of fault response under multi-source uncertainty was quantified. Isolability is defined based on the fault space and transformed into the inherent properties of the corresponding fault space, meanwhile, quantification of fault diagnosability is transformed into solving multivariate distribution probability distance, and the evaluation of detection and isolation is realized by introducing Bhattacharyya distance. Numerical simulation results show that the diagnosability evaluation method proposed in this paper is effective under multi-source uncertainty and can realize the quantitative evaluation of the detectability and isolability dynamically during the fault response.
AB - Aero-engine is faced with various uncertainties while operating. To reduce the conservativeness of diagnosis system in complex environment, it is of great significance to consider the impact of multi-source uncertainty of fault diagnosis systems during the design process, the most effective way of which is to quantify the fault diagnosability. In this paper, the multi-source uncertainty environment, which is composed of aero-engine epistemic uncertainty caused by modeling and stochastic uncertainties of actuator output and sensor measurement in control system, is considered, and the diagnostic evaluation of actuator, sensor and gas-path component faults is carried out, respectively. Based on ITO integral and polynomial chaos expansion, the agent model of output response was established, and the uncertainty of fault response under multi-source uncertainty was quantified. Isolability is defined based on the fault space and transformed into the inherent properties of the corresponding fault space, meanwhile, quantification of fault diagnosability is transformed into solving multivariate distribution probability distance, and the evaluation of detection and isolation is realized by introducing Bhattacharyya distance. Numerical simulation results show that the diagnosability evaluation method proposed in this paper is effective under multi-source uncertainty and can realize the quantitative evaluation of the detectability and isolability dynamically during the fault response.
KW - Aero-engine
KW - Bhattacharyya distance (BD)
KW - Fault diagnosability
KW - Multi-source uncertainty
KW - Polynomial chaos expansion (PCE)
KW - Uncertainty quantification (UQ)
UR - http://www.scopus.com/inward/record.url?scp=105004455424&partnerID=8YFLogxK
U2 - 10.1007/s42405-025-00944-4
DO - 10.1007/s42405-025-00944-4
M3 - 文章
AN - SCOPUS:105004455424
SN - 2093-274X
JO - International Journal of Aeronautical and Space Sciences
JF - International Journal of Aeronautical and Space Sciences
M1 - 108058
ER -