Distributed optimization for uncertain Euler–Lagrange Systems with local and relative measurements

This paper considers a multi-agent system modeled as a group of Euler–Lagrange systems, and assumes that each agent only has perception of the real-time position-dependent gradient value of a local objective function and its relative position with other agents. Based on a seamless integration of a modified distributed optimization algorithm and a Lyapunov-based nonlinear control design, distributed controllers are developed to exponentially steer the position of each agent to the optimal point of the total objective function. A numerical example of coordinated communication relay is employed to verify the proposed design.