Consensus of complex networks based on the agents´ output

This paper addresses the consensus problem of the multi-agent systems with directed communication topology consisting of general linear node dynamics. By some similar transformation and the augment of the output matrix, the consensus of the multi-agent systems based on the agents´ output can be transformed into the constrained stability of some systems based on the system state. Consensus problem can then be cast into the feasibility of a series of linear matrix inequalities(LMIs) with the complex eigenvalue as the coefficient, which can be converted to the real coefficient linear matrix inequalities through the equivalence of some matrix inequalities. This method is also applied to tracking a reference state. Some numerical examples are presented to illustrate the effectiveness of the proposed method.