Decentralized Control of Networked Systems with Unknown Control Directions and Uncertainties

In this paper, we investigate the decentralized control of networked agents with unknown control directions, i.e., the sign of high-frequency-gain in agent's dynamics is unknown. Meanwhile, uncertain nonlinearity also appears in agent's dynamics. A Nussbaum-type adaptive controller is designed with the aid of nonlinear damping technique for each agent such that consensus of the network can be achieved while all signals in the overall system maintain bounded. The controller for each agent is composed by a Nussbaum-type item which seeks the control direction adaptively and a distributed item to drive the system to consensus state, in which a nonlinear damping item is added. It is proved that consensus of the network can be realized via the proposed controller. Moreover, the final consensus state is indicated, which is significant for consensus of networked nonlinear systems. Simulation results are presented to verify the effectiveness of the proposed controllers.