A control method based on meliorated Potential Fields for dynamic coverage problem with multi-agents

With the development of UAV technology, more and more attentions have been paid to control multi-agents to solve a huge problem. The Potential Fields is one of the most intelligent control methods. It moves agents intuitively by creating virtual forces depending on available information. In this paper, some meliorations were made on the Potential Fields to meet the demands of a dynamic coverage problem with multi-agents. Virtual forces helped the agents avoid collision and maintain connectivity while the flocking was covering an undiscovered area under the guidance of a brief digital map. The variable force principle was brought in to help optimize the distribution of agents automatically. The discarding attraction principle was adopted to resist local minima. Finally, most of the area was successfully covered in simulations, when only local wireless communication is available.