Distributed Optimization for Nonlinear Multi-Agent Systems with Upper-Triangular Structure

Abstract

The distributed optimization for nonlinear multi-agent systems in the form of upper-triangular is considered in this paper. To handle this issue, a new two-step framework is proposed to derive distributed controllers, under which the system outputs are consistent with the optimal solution. Using the real-time values of gradient functions, the distributed optimal coordinators are established to produce signals approaching the optimal solution. Then, by developing the feedback domination approach, the reference-tracking controllers are designed with the help of a low-gain parameter, which can globally asymptotically stabilize the nonlinear multi-agent systems. Finally, an example is simulated to illustrate the main idea.