Flexible Skin for Flight Parameter Estimation Based on Pressure and Velocity Data Fusion

Abstract

Accurate perception of flight parameters is essential for the safe and stable flight of small unmanned aerial systems (SUASs). However, traditional flush air data sensing (FADS) systems require complex tubing and fuselage openings. Herein this study, a flexible skin system integrated with a pressure and thermal flow sensor array, and a dual‐sensor fusion algorithm for determining the angle of attack (AOA) and airspeed (V ∞ ), are proposed. By establishing an error back‐propagation neural network, the dual‐sensor fusion modality demonstrates higher estimation accuracy than other modalities that utilize only pressure data or only flow velocity data, achieving mean absolute errors of the AOA and V ∞ of less than 0.16° and 0.37 ms−1, respectively. Moreover, the simulation and experimental results show that sensors placed closer to the leading edge of the wing can provide higher estimation accuracy of the flight parameters. The flight parameters determined using the flexible air data sensing system and dual‐sensor fusion algorithm demonstrate potential application in the flight control of SUASs.

Keywords

• angle of attack
• flexible skin
• flow sensor
• neural network
• sensor fusion