Robust Design Optimization of Unmanned Aerial Vehicle Coaxial Rotor Considering Operational Uncertainty

This research aims at performing robust design optimization by taking into account operational uncertainties to improve the performance of the coaxial rotor unmanned aerial vehicle that is under development. For this sake, neural network models are constructed for aerodynamic performance obtained from a source-doublet panel method coupled with a time-marching free wake method. Two kinds of operational uncertainties are considered: ballistic damage on blades and weight variation due to the mission change. Approximate moment approach is employed to evaluate the robustness of performance. Subsequently, a relationship between the robustness of performance indices and the parameters representing blade geometry is investigated. In addition, the robustness of the performance indices and constraints for three different solutions chosen from the Pareto optima set is studied. It is confirmed that the trend of design variables to improve performance is in contradiction to the robustness of design. Besides, by incorporating the robust design method, a probability for a designed coaxial rotor to accomplish missions successfully can be improved by 70 similar to 95% under the uncertainty of ballistic damage.