Optimized observation arrangement based on multi-agent

In the configuration that a formation is chased by a group of chasing vehicles, it is a common strategy for the formation to observe the chasing vehicles coordinately. Collaborative observation assignment need to be optimized to maximum the ability and effectiveness of observation with minimum switching cost. Master-Slave scheme relies on the center node, so the reliability and robust of formation is affected. In this paper, a multi-agent algorithm based observation arrangement optimizing strategy is proposed. In the proposed strategy, each member of formation acts as agent separately and collaboratively. In each agent, from the guidance recognition results of chasing vehicles, the chased target member is identified and overall threat level of chasing group is evaluated. Rearrangement decision or the switch of observation target, is made individually on each agent's local observation and shared information. Simulation results proved that with distributed strategy deployed on individual members of a formation, the overall threat level of chasing group can be minimized with optimized switching cost. Even in extreme conditions that some members failed, the observation ability can be maintained at acceptable level.