Flapping Wing UAV Neural Network Model

Project Description

This project involves training a Multi-Layer Perceptron (MLP) model on flapping wing UAV flight data. The model predicts aerodynamic characteristics (lift and induced drag) based on design parameters and flight conditions.

Input Parameters

• Airfoil type
• Wingspan
• Taper ratio
• Aspect ratio
• Angle of attack
• Airspeed
• Flapping period

Output Parameters

• Lift
• Induced drag

Requirements

Core Dependencies

• Python 3.8
• CUDA 11.8
• PyTorch 2.4.1
• NVIDIA GPU with CUDA support

Installation

1. Clone the repository:

git clone https://github.com/Srindot/Deepnn-Average-Flight-Data-for-FWUAV.git
cd flapping-wing-uav-nn

2. Create and activate conda environment:

conda create -n flap-uav python=3.8
conda activate flap-uav

3. Install dependencies:

pip install -r requirements.txt

Usage

1. Data preparation is in data_collection.py
2. Model training is in train_model.py
3. Trained model weights are saved in .pth format

Model Architecture

• Multi-Layer Perceptron (MLP)
• Input features: Design and flight parameters
• Output: Lift and induced drag predictions