Event-Driven Robust Guaranteed Cost Control via an Improved Adaptive Critic Learning Strategy

In this paper, we develop an event-driven robust guaranteed cost control strategy of continuous-time (CT) systems via improved adaptive critic learning (ACL). First, we choose a suitable cost function which reflects uncertainties, control, and regulation, in order to transform the robust control problem into the optimal control problem. Then, we obtain the time-driven optimal control law and the Hamilton-Jacobi-Bellman equation. Next, through theoretical analysis, we derive the event-driven optimal control law of the nominal system based on the ACL method, and prove the robust stabilization of the CT nonlinear system. Additionally, we construct a novel critic neural network learning algorithm to accelerate the convergence of weights. We also obtain the neural-network-based event-driven condition and prove the closed-loop system stability. Finally, the simulation result shows the effectiveness of the event-driven guaranteed cost control design.