Electronic Thesis/Dissertation
 

A Practical Bayesian Approach for Improved Monitoring of Aircraft Operational Performance

Open Access Deposited

Fuel is a major operating expense for airlines, which use specialized Aircraft Performance Monitoring (APM) systems to track aircraft fuel consumption trends. However, legacy monitoring systems cannot accurately detect variations in aircraft fuel usage smaller than 1% or reliably distinguish between airframe and engine deterioration. While recent literature presents machine learning models as alternatives, these approaches sacrifice interpretability without demonstrating better accuracy than physics-based approaches, limiting their diagnostic utility for maintenance interventions.This dissertation introduces a hierarchical Bayesian model that combines physics-based models, parametric relationships, and domain expertise to separately identify and quantify aerodynamic and propulsive deterioration that impacts fuel consumption on individual aircraft in an operational fleet. A two-tiered approach first establishes a baseline physical model with fleet-wide parameter estimates, then models individual aircraft-specific deviations from the baseline with credible intervals. Verification using a simulated dataset with known airframe and engine deterioration in a virtual fleet of 35 Boeing 737-800 aircraft achieves better than 1% accuracy, with mean absolute errors of 0.31% for aircraft-specific drag increments and 0.21% for engine-specific fuel-flow parameters. Validation with NASA-supplied data from a fleet of 35 RJ85 regional jets yields consistent results, with per-engine fuel-flow residuals of 0.69% to 0.71%. The validated methodology successfully detects sub-1% performance variations with quantified uncertainties and distinguishes airframe drag effects from engine-specific deterioration, enabling targeted maintenance interventions unavailable in legacy performance monitoring approaches.

Author Language Keyword Date created Type of Work License
  • All rights reserved
Rights statement GW Unit Degree Advisor Committee Member(s) Persistent URL

Notice to Authors

If you are the author of this work and you have any questions about the information on this page, please use the Contact form to get in touch with us.

Thumbnail Title Date Uploaded Visibility Actions
Preview of Bays_gwu_0075A_17700.pdf Bays_gwu_0075A_17700.pdf 2025-12-15 Open Access