Distributed Adaptive Consensus of Nonlinear Heterogeneous Agents With Delayed and Sampled Neighbor Measurements

In this article, the adaptive output consensus problem of high-order nonlinear heterogeneous agents is addressed using only delayed, sampled neighbor output measurements. A class of auxiliary variables is introduced which are n-times differentiable functions and include the agent's output along with delayed, sampled output neighbor measurements. It is proven that if these variables are bounded and regulated to zero then asymptotic consensus among all agent outputs is ensured. In view of this property, an adaptive distributed backstepping design procedure is presented that guarantees boundedness and regulation of the proposed variables. This design procedure ensures not only the desired asymptotic output consensus but also the uniform boundedness of all closed-loop variables. The main feature of our approach is that, in the proposed control law for each agent, the entire state vector of the neighbors is not needed and only delayed sampled measurements of the neighbors' outputs are utilized. The simulation results are also presented that verify our theoretical analysis.