Regime Recognition (RR) is a key enabling technology for a Usage Based Maintenance (UBM) program. If the estimated aircraft usage spectrum provided by such an RR algorithm is sufficiently accurate, then component life expenditure can be calculated as a function of the duration and number of occurrences of each flight regime. In general, production RR algorithms with the required accuracy have not been fielded, in part because of the challenges associated with Verification and Validation (V&V). In particular, the distinction between transient (event-based) regimes and steady-state (duration-based) regimes is an important consideration when defining performance metrics in a V&V process. While duration-based regimes lend themselves well to traditional analysis metrics that employ a confusion matrix, event-based regimes lack the error properties that make such analysis methods possible. In this paper we present analysis methods that address this deficiency and allow the use of well understood metrics derived from the traditional confusion matrix.
Regime recognition (RR) enables indirect characterization of high and low cycle fatigue damage to critical aircraft components during fleet usage based on structural maneuvers and measured loads from a flight test program. To ensure that an RR algorithm is suitable for usage or condition-based maintenance (UBM/CBM), it is arguably necessary to guarantee that it performs consistently during the design phase, when presented with structural maneuvers precisely executed by test pilots during a flight test program, and during deployment, when presented with general maneuvers flown by fleet pilots that include damaging structural maneuvers or something similar enough to be considered structural maneuvers. Dynamic time warping (DTW) is used to provide a framework to efficiently and automatically identify flight segments in fleet data that resemble flight-test structural maneuvers for which component loads are known; a normalized metric based on DTW distance measures is proposed to quantify the similarity and control the procedure to pull similar maneuvers from the fleet data. The resultant distribution of this metric is used to quantify variability of a structural maneuver of interest and in turn provide guidance on the number of unique maneuvers that should be used in a verification and validation (V&V) process to analyze a candidate RR algorithm.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.