Recalibration of Flush Air Data System (FADS) on Pressure Port/Sensor Failures

Estimation of Airflow Parameters for Tail-Sitter UAV through a 5-Hole Probe Based on an ANN

“…From above figures and analysis, we can draw the following conclusions: When the angle of attack and sideslip angle are not zero at the same time, the neural network method can still well predict the angle of attack, sideslip angle and velocity of incoming flow. It can be seen in Figure 14 and Figure 15 that when the angle of attack gradually increases to about 50°, the estimation of the neural network method can still remain within ±1°, whereas the absolute values of the estimation error of other methods introduced in the first part exceed 1° or even more in this case [ 5 , 6 , 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 , 25 ]. This is because the combination of pressure characteristics brought about by the five pressure-measuring holes in different directions can better correspond to this specific combination of flow parameters.…”

“…The algebraic solution is usually simple. The least square method and three-point method [ 11 , 12 , 13 , 14 , 15 , 16 ] establish the aerodynamic model of the probe surface, obtain the key parameters through the experiment and then transform the prediction problem into a simple problem that can be solved algebraically. However, in order to improve the accuracy of these models, they not only need more experiments, but also need to establish the relevant error model.…”