Continuous patrolling in uncertain environment with the UAV swarm

Area of Science:

• Robotics and Control Systems
• Artificial Intelligence
• Distributed Systems

Background:

• Unmanned Aerial Vehicle (UAV) swarms with sensors are increasingly used for situational awareness.
• Current research inadequately addresses complex control structures in UAV swarms.
• Dynamic and uncertain environments pose challenges for UAV swarm operations.

Purpose of the Study:

• To propose a novel patrolling task planning algorithm for UAV swarms with a double-layer centralized control structure.
• To maximize environment information collection by the UAV swarm.
• To address the limitations in current research on complex UAV swarm control.

Main Methods:

• Defined the patrolling problem for UAV swarms.
• Modeled the patrolling problem as a Partially Observable Markov Decision Process (POMDP).
• Developed a myopic and scalable online task planning algorithm incorporating heuristic functions, sequential allocation, and information/command flow mechanisms.

Main Results:

• The proposed algorithm effectively reduces computation complexity.
• Performance is maintained without increasing computation complexity for the swarm leader, even with more control layers.
• Empirical evaluation in specific scenarios demonstrated the algorithm's efficacy.

Conclusions:

• The developed algorithm offers an effective solution for UAV swarm patrolling in uncertain environments.
• It provides a scalable and computationally efficient approach to task planning for complex UAV swarms.
• The algorithm enhances the capability of UAV swarms to achieve situational awareness through effective information gathering.