LAUSR

This paper studies an optimal control problem for multi-agent systems under adaptive dynamic programming (ADP) framework. To overcome the restrictions resulting from the limited calculation capabilities of agents and to prolong the lifetime of actuators, an event-triggered mechanism is first considered. Different from the traditional actor-critic dual networks architecture, a simplified ADP is developed. In addition, the weights in the critic network are updated based on the gradient descent method and the experience replay technique, and thus, the persistence of excitation condition is no longer needed. It is proved that all signals in the resulting closed-loop systems are uniformly ultimately bounded via the Lyapunov analysis. Finally, simulation examples are provided to illustrate the effectiveness of the proposed method.