Virtual formations and virtual leaders in formation control problem for group of robots

This article considers important problem of robotics — formation control for autonomous agents with decentralised control approach. The agents must build a specific formation and move towards the target point maintaining it. The suggested method is based on virtual formations and virtual leaders. The rule for calculating the position of virtual leaders and their priorities for agents is introduced. Only local information is used in suggested control rule. The position of the agent within the formation is chosen dynamically, the agents dynamically react on the change of target point position and calculate the trajectory by themselves. The suggested approach doesn't require sending to agents neither the information about the current coordinates nor the information about the direction of the virtual formation and virtual leaders. The geometric structure of formation switches in case of agent failure or in case of establishing connection with another agent, agents are fully interchangeable. The control rule is proven by computer modelling results. Bibliogr. 13. Il. 4.