Neural Network Based Predictions for the Aerodynamic Performance of Flapping Wings

There is unrealized potential in using Neural Network approaches in aerospace, particularly in assisting with aircraft design. The objective of this thesis was to determine if neural networks could predict the morphology, kinematics, and performance of flapping wings. A neural network was developed in Python and trained using a small dataset of biological insect data from literature. The model was then tested for biological data (small and large insects) as well as micro-aerial vehicles. The small insect test dataset performed best likely due to similarities with the training set. A larger dataset is needed to validate the use of neural networks in flapping wing micro-aerial vehicle design. This work is ground-breaking in the field of flapping wing micro-aerial vehicle design since it provides the foundation for a quick and accurate alternative to lengthy experiments and simulations.