Event-Triggered Optimal Attitude Consensus of Multiple Rigid Body Networks With Unknown Dynamics

In this paper, an event-triggered Reinforcement Learning (RL) method is proposed for the optimal attitude consensus of multiple rigid body networks with unknown dynamics. Firstly, error constructed through According to Bellman optimality principle, implicit form controller and corresponding Hamilton-Jacobi-Bellman (HJB) equations are obtained. Because augmented system, can be obtained directly without relying on system Secondly, self-triggered mechanism applied reduce computing communication burden when updating controller. order address problem that HJB difficult solve analytically, RL which only requires measurement data at instants proposed. For each agent, one neural network designed approximate value function. Each updated instants. Meanwhile, Uniformly Ultimately Bounded (UUB) closed-loop obtained, Zeno behavior also avoided. Finally, simulation results a demonstrate validity method.