Large Angle of Attack Prediction for Tail-Sitter using ANN-based Flush Air Data Sensing

ABSTRACT

Flush air data sensing systems (FADS) have been widely applied on aerial vehicles to provide air data estimation. Air data such as angle of attack (AoA) and air speed can be estimated through resolving pressure measurements of the sensor matrix. These parameters can be utilized to improve the performance of flight control system and realize better flight performance. Existing FADS studies and applications can estimate AoA in a range typically below 55 degree. It is suitable for traditional fixed wing unmanned aerial vehicles (UAVs), but some fixed wing vertical take off and landing (VTOL) UAVs have requirements in measuring air data under larger AoA. In this work, a FADS based on artificial neural network have been applied on a tail-sitter to provided large AoA estimation in low Reynolds number. Computational fluid dynamic analysis have been carried out to evaluate the critical AoA where stall region affects the sensor matrix. Wind tunnel tests have been further carried to collect data for network training. The trained network can provide estimation of large AoA at the range of -80 degree to 80 degree with acceptable accuracy.

Mr. Liao Yangtianchun is pursuing his PhD study in the Collaborative PhD Program of National University of Singapore and Southern University of Science and Technology. His research interest includes high angle of attack air data sensing and physics informed neural networks.