Decentralized Coordination of a Multi-UAV System for Spatial Planar Shape Formations.

Area of Science:

• Robotics
• Artificial Intelligence
• Computer Science

Background:

• Firefighters require effective methods for monitoring hazardous environments like toxic clouds.
• Existing drone coordination methods may lack decentralization and adaptability for complex tasks.

Purpose of the Study:

• To develop a decentralized technique for a fleet of drones to collectively form arbitrary planar virtual shapes in 3D space.
• To enable drones to surround and monitor hazardous areas, such as toxic clouds or fire perimeters.

Main Methods:

• Utilized Fourier descriptors for precise mathematical description of virtual shapes.
• Implemented a decentralized bearing angle assignment for even drone distribution along the shape.
• Developed algorithms for drone movement, collision avoidance, and dynamic fleet size adjustment.

Main Results:

• The proposed method successfully enables decentralized virtual planar shape description and drone assignment.
• Demonstrated drone movement from takeoff to assigned positions along the shape with collision avoidance.
• Validated the technique's applicability in real-world scenarios through simulations, emulation, and outdoor experiments with real drones.

Conclusions:

• The developed decentralized drone formation technique is effective and adaptable for real-world applications.
• This method enhances drone capabilities for environmental monitoring and emergency response.
• The technique supports dynamic changes in the number of drones during a mission.