This project, undertaken at IIT Bombay under the guidance of Prof. Avinash Bharadwaj, focuses on optimizing Unmanned Aerial Vehicle (UAV) networks for rescue operations in non-convex domains. The work involves:
- Mathematical Formulation: The Search and Rescue (SaR) problem is mathematically formulated as an optimization problem to minimize time.
- Genetic Algorithm: A novel genetic algorithm-based method is implemented to maximize the area of coverage for UAV surveillance tasks in non-convex and disconnected placement domains.
- Image Processing Framework: An image processing framework, utilizing the Ramer-Douglas-Peucker algorithm, is devised to approximate region boundaries to polygonal sets, ensuring the autonomous functioning of the algorithm.
The results from this project were presented in the paper:
Arpit Dwivedi, Chinmay Pimpalkhare. "Coverage Maximization for UAV Surveillance on Non-convex Domains using Genetic Algorithm." Paper presented virtually at the 6th National Conference on Multidisciplinary Design, Analysis and Optimization, IIT Guwahati, India.
The repository contains the following main folders and files:
- Genetic Algo: Contains the implementation of the genetic algorithm.
- Images: Contains image data used for processing.
- Python Files: Contains various Python scripts related to the project.
To run the pipeline, use the following command in your terminal:
python3 Rescue_optimization.py --calamity <Calamity_type> --rescuerobots <Number_of_rescue_robots> --typerr <Types_of_Robots> --sensorrobots <Number_of_sensor_robots> --typesr <types_of_sensor_robots>
You can adjust the placeholders and content to match the specific details of your project.
In order to run the pipeline, write the following in terminal:
python3 Rescue_optimization.py --calamity <Calamity_type> --rescuerobots <Number_of_rescue_robots>> --typerr <Types_of_Robots> --sensorrobots <Number_of_sensor_robots> --typesr <types_of_sensor_robots>