Neural-network-based adaptive guaranteed cost control of nonlinear dynamical systems with matched uncertainties

In this paper, we investigate the neural-network-based adaptive guaranteed cost control for continuous-time affine nonlinear systems with dynamical uncertainties. Through theoretical analysis, the guaranteed cost control problem is transformed into designing an optimal controller of the associated nominal system with a newly defined cost function. The approach of adaptive dynamic programming (ADP) is involved to implement the guaranteed cost control strategy with the neural network approximation. The stability of the closed-loop system with the guaranteed cost control law, the convergence of the critic network weights and the approximate boundary of the guaranteed cost control law are all analyzed. Two simulation examples have been conducted and all simulation results have indicated the good performance of the developed guaranteed cost control strategy.