Using Indirect Turbulence Measurements for Real-Time Parameter Estimation in Turbulent Air

Abstract

The use of indirect turbulence measurements for real-time estimation of parameters in a linear longitudinal dynamics model in atmospheric turbulence was studied. It is shown that measuring the atmospheric turbulence makes it possible to treat the turbulence as a measured explanatory variable in the parameter estimation problem. Commercial off-theshelf sensors were researched and evaluated, then compared to air data booms. Sources of colored noise in the explanatory variables resulting from typical turbulence measurement techniques were identified and studied. A major source of colored noise in the explanatory variables was identified as frequency dependent upwash and time delay. The resulting upwash and time delay corrections were analyzed and compared to previous time shift dynamic modeling research. Simulation data as well as flight test data in atmospheric turbulence were used to verify the time delay behavior. Recommendations are given for follow on flight research and instrumentation.

Aircraft Dynamic Modeling in Turbulence

“…This approach allowed in-flight turbulence variation to be automatically tracked. Because the flights were all short (approximately 15 minutes) and in general had little temporal or altitude variation, the mean value of w [2][3][4][5][6][7] for the entire research segment of the flight was used to compare with pilot comments. Very good agreement and sensitivity is shown in Figure 4, which illustrates the correlation of the turbulence characterization metric with the pilot's subjective turbulence rating.…”

“…For this analysis, the metric, w [2][3][4][5][6][7] , was computed at 1 second intervals based on a sliding 30-second window of data. This approach allowed in-flight turbulence variation to be automatically tracked.…”

“…The equation-error method in the frequency domain is particularly well-suited to this task, and can be implemented as a real-time method 1 . Separate work 4 has studied the possible use of turbulence sensors to convert the turbulence from an unknown excitation input to a known excitation input, after which standard parameter estimation methods could be applied. In that case, the turbulence is being incorporated into the model, but using a measurement rather than an analytical estimation technique.…”

Validation of Safety-Critical Systems for Aircraft Loss-of-Control Prevention and Recovery

“…42 Recent and ongoing research into icing effects focuses on airframe icing 43 as well as engine icing. [44][45] The modeling and detection of airframe icing effects under turbulence conditions is also the subject of current research, [46][47] as is the modeling and detection of engine icing effects. Simulation-based evaluation methods are also being used to evaluate the potential effectiveness of integrated flight and enhanced propulsion control capability for LOC prevention and recovery, 49 as well as for determining the achievable dynamics of an impaired vehicle.…”

Loss of Control Prevention and Recovery: Onboard Guidance, Control, and Systems Technologies

“…9, much of this information would be provided to the crew for improved SA and anticipatory guidance. Research is currently being conducted in icing effects detection and identification in the presence of turbulence, [14][15] engine icing effects modeling and detection, 16 energy management, 17 LOC prediction, 18 and improved SA [19][20] . Resilient control and anticipatory guidance methods are also of key importance in preventing LOC accidents, and are indicated by the middle two subsystems of Figure 9.…”