A real-time method for estimating time-varying aircraft frequency responses from input and output measurements was demonstrated. The Bat-4 subscale airplane was used with NASA Langley Research Center's AirSTAR unmanned aerial flight test facility to conduct flight tests and collect data for dynamic modeling. Orthogonal phase-optimized multisine inputs, summed with pilot stick and pedal inputs, were used to excite the responses. The aircraft was tested in its normal configuration and with emulated failures, which included a stuck left ruddervator and an increased command path latency. No prior knowledge of a dynamic model was used or available for the estimation. The longitudinal short period dynamics were investigated in this work. Time-varying frequency responses and stability margins were tracked well using a 20 second sliding window of data, as compared to a post-flight analysis using output error parameter estimation and a low-order equivalent system model. This method could be used in a real-time fault detection system, or for other applications of dynamic modeling such as real-time verification of stability margins during envelope expansion tests.
Aircraft Fault Detection Using Real-Time Frequency Response Estimation
2016
19 pages
Report
Keine Angabe
Englisch
Transportation Safety , Air Transportation , Flight tests , Fault detection , Real time operation , Frequency response , Estimating , Dynamic models , Time , Variations , Error analysis , Aircraft maneuvers , Postflight analysis , Parameter identification , Mathematical models , Pedals , Test facilities , Remotely piloted vehicles , Control surfaces , Failure , Stability
Aircraft Fault Detection Using Real-Time Frequency Response Estimation (AIAA 2016-0372)
British Library Conference Proceedings | 2016
|Real-time monitoring system for aircraft fault detection
Europäisches Patentamt | 2021
|