Bipartite hybrid formation tracking control for heterogeneous multi-agent systems in multi-group cooperative-competitive networks

Abstract

In this paper, the problem of bipartite hybrid formation tracking for multi-agent systems with heterogeneous groups is studied. First, the network structure problem of a cluster of agents with both group relationship and cooperation-competition relationship is discussed. Then, based on the leading-following method, the system model consisting of nonlinear second-order dynamic agents and first-order linear dynamic agents is determined. In this network system, the first group of second-order dynamic individuals needs to track their leader with unknown input and realizes the bipartite time-invariant formation, while the other group tracks its leader to realize the bipartite time-varying formation. To solve this problem, a new control protocol based on neighbor information and leader dynamics is designed. In addition, for first-order dynamic individual, a velocity estimator is introduced to estimate its velocity state. Next, algebraic graph theory and Lyapunov stability theory are used to prove that the formation error can approach arbitrary neighborhood of zero. Finally, the simulation results show the feasibility of the proposed conclusions and the effectiveness of the designed velocity estimator.