Self-Triggered Robust Optimal Consensus Control for Nonlinear Multi-Agent Systems with Control Constraints

This paper proposes a self-triggered adaptive dynamic programming (ADP) approach for robust optimal consensus control of nonlinear multi-agent systems (MASs) with input constraints. Unmatched uncertainties are transformed into an optimal control framework via augmented techniques, and a nonquadratic cost function is used to handle actuator saturations. To reduce communication and computation overhead, an integrated self-triggered mechanism is developed. A critic-only neural network approximates the optimal solution. Lyapunov analysis proves closed-loop robust stability and uniform ultimate boundedness (UUB) of weight estimation errors. Simulation results validate the effectiveness of the proposed method.