Convergence speed in formation control of multi-agent systems - A robust control approach

In this work a robust design approach to improve the convergence speed in formation control of multi-agent systems by simultaneously shaping the control feedback law and the weights of the interconnection structure is proposed. The communication topology is assumed to be fixed and undirected. It is shown that for agents modeled as discrete-time double-integrators, convergence speed of the proposed method almost reaches the theoretical optimum. The convergence speed can be further increased if the feedback control law is allowed to be dynamic. Although the number of decision variables in this approach grows linearly with the size of the weighting matrix, the synthesis method is applied also to a more realistic linearized model of a quad-rotor helicopter, where the simulation results indicate that this approach is reasonable for a sufficiently large number of high-order agents.