Bad weather conditions during an air flight (E.g. Hailstorms) can cause extensive damage to aircraft structure and compromise its structural integrity. Predicting the sudden impact of hailstorms is a major problem in the aviation industry. Aim of this experiment is to predict the impact location and/or its characteristics in a controlled experimental setup. The output of the neural network should predict the location of impact of the ball, based on the piezoelectric sensor data as input.
For a more detailed explaination of the code and results please refer to the attached PPT.
As an addendum to this project, a novel approach using artificial augmentation of the data is also implemented. For more details, refer to the Addendum Report.