Research On Region-based Formation Control For Multi-agent Systems

Now,with the increasing application of UAV,UGV,etc.in cooperative work,formation control of multi-agent systems is more valuable.For shape control,we focus on the region-based control,with consideration of its first-order and second-order control models,the formation characteristics,and relationship with flocking and area coverage.First,I built the first-order model of the region-based formation control for multiagent systems,and introduced two artificial potential fields constructed in this method:the global artificial potential field(GAPF)attracted by the region and the local artificial potential field(LAPF)between agents.With regard to whether the desired area is related to states of agents,I considered two types: environmental(agent-independent)areas and virtual(agent-dependent)areas,and proved the convergence of systems in these cases,along with comparisons of the difference in steady state,particularly it is time-varying and collective when the geometric center of the system is used as the reference point of the target area,which will make the speed of the stable state be uncertain.Then,considering the multi-agent second-order model,I combined the flocking behavior and the region-based formation control,incorporating the distributed consensus idea including position coupling and velocity coupling into the regional formation control,to achieve the goal of forming the shape and having the same speed simultaneously.I constructed a Lyapunov function to prove the convergence of the system under this hybrid controller.In addition,I also considered the complex shape formation control.Last,I found and defined two modes of multi-agent stable states with region-based formation control: interior-regional and non-interior-regional formation.I qualitatively and quantitatively analyzed how the two artificial potential fields influence the mode.Then,I proposed an adaptive interior-regional controller to make the system converge to the interior-regional formation.Further,combining the perturbation method,I proposed an area-coverage algorithm.