In this paper, an event-triggered distributed ${ {\mathcal {H}}_{\infty }}$ constrained control algorithm is designed for physically interconnected large-scale partially unknown strict-feedback systems with constrained-input and external disturbance. The advantage… Click to show full abstract
In this paper, an event-triggered distributed ${ {\mathcal {H}}_{\infty }}$ constrained control algorithm is designed for physically interconnected large-scale partially unknown strict-feedback systems with constrained-input and external disturbance. The advantage of both physical interconnection and communication is synchronously exploited for the scheme. First, an event-triggered feedforward control policy is proposed to transform control of physically interconnected large-scale systems into equivalent event-triggered control of decoupled multiagent systems. Then, an event-triggering condition and an event-triggered feedback control algorithm are designed to learn the optimal control policy and the disturbance policy in the worst case. The algorithm eliminates identifier, actor, and disturber neural networks and also relaxes the persistent excitation condition. It guarantees that the closed-loop dynamics is stabilized and the cost function is converged to the bounded ${\mathcal {L}}_{2}$ -gain optimal value while the Zeno phenomenon is excluded. Finally, the effectiveness of the proposed algorithm is verified through simulation results of event-triggered distributed control of a physically interconnected constrained-torques multimobile robot system.
