Abstract
Aircraft agility metrics have been proposed for use to measure the performance and capability of aircraft onboard while in-flight. The sensitivity of these metrics to various types of errors and uncertainties is not understood. This research effort develops a framework for analyzing the sensitivity of agility metrics to initial condition errors, structured parametric uncertainties, and coordinate system type without the need to resort to time consuming and costly Monte Carlo analysis. Linear error theory is used to quantify the sensitivity of a selected set of agility metrics, and the results are then verified with both time histories and traditional Monte Carlo methods. The four selected metrics are time to roll through bank angle metric, time-averaged integral of pitch rate metric, power onset/loss parameter metric, and the Combat Cycle Time metric. Results show that the time to roll through bank angle metric is weakly nonlinear to initial condition errors and parametric uncertainties. The time-averaged integral of pitch rate metric is linear to initial condition error in pitch rate. The power onset/loss parameter metrics are strongly nonlinear to initial condition errors and parametric uncertainties, and the Combat Cycle Time metric is also strongly nonlinear to initial condition errors and parametric uncertainties. Using the Beck metrics expressed in the Frenet coordinate system, the time to roll through bank angle metric is more sensitive to initial condition errors and parametric uncertainties than when expressed in the Cartesian coordinate system. The time-averaged integral of pitch rate and power onset/loss parameter metrics were less sensitive to initial condition errors and parametric uncertainties when expressed using the Beck metrics than when expressed in the Cartesian coordinate system. The Combat Cycle Time metric results indicate little difference between the sensitivity of the metrics to coordinate system type. Finally, the linear error theory method for analyzing the sensitivity of agility metrics has been verified.
Smith, David Matthew (2000). Agility metric sensitivity using linear error theory. Master's thesis, Texas A&M University. Available electronically from
https : / /hdl .handle .net /1969 .1 /ETD -TAMU -2000 -THESIS -S613.