Model Identification of Quadrotor Aerodynamic Interactions in Forward Flight
Loading...
Date
2024-05
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
The Ohio State University
Abstract
Multirotors exhibit many unique flight characteristics including hovering, vertical taking-off and landing, and high maneuverability. These unique features have attracted many industries to utilize multirotors to help enhance the transportation, mapping, surveillance, and aerial photography industries. However, complex aerodynamic interactions including rotor-rotor and rotor-airframe cause challenges in developing a high-fidelity model over the entire flight envelope.
The primary objective of this work was to develop a higher fidelity system identification model using open-air flight test data, which fully encompasses the interactions present in trimmed flight. A gray box modeling approach was used, which combined the limited prior knowledge of vehicle dynamics and open-air flight test data in terms of the rotor forces and moments. However, some parameters remain unknown including rotor-rotor and rotor-airframe interactions. In addition, rotor geometry estimation was an important input to the model, which required chord and pitch data along the radial span of the flight test rotor. XFOIL was used to obtain the aerodynamic coefficients of the blade.
Results from incorporating the system identification model to the theoretical model showed improved accuracy when compared to flight test data. Each regressor selected in the least squares model was significant as the r-squared value for each rotor was greater than 0.85. After completing the rotor geometry estimation using visual imaging, a piecewise function was identified capturing the change in chord and pitch along the radius of the rotor. Also, it was found that the airfoil best represents a GOE 195 with slight rounding modifications on the leading and trailing edge for ease of manufacturing.
Overall, this work supports a larger project focused on studying the influence of multirotor interaction effects using a linear inflow model based on Blade Element Theory (BET). The pitch and chord functions found were utilized in the model to predict isolated rotor thrust and torque for the propulsion system used on the flight-testing vehicle. This application of system identification was proved valid and future work requires adding the aerodynamic interaction model into the flight controller and testing the quadrotor’s performance in trimmed flight. Results from the project will advance the accuracy of future system models to aid in reliable controls system design, risk analysis, and flight simulators.
Description
Keywords
System Identification