Robust adaptive quadratic tracking control of continuous-time linear systems with unknown dynamics

In this paper, an online robust adaptive quadratic tracking control algorithm is proposed for continuous-time linear systems with unknown dynamics and nonlinear dynamic uncertainties. The robust stability and suboptimality of the closed-loop system composed of the considered system and the quadratic tracking policy is proved. For finding the solution of the linear quadratic tracking (LQT) problem without knowing any know-ledge of the system matrices and uncertainties, a new policy iteration approach is presented. The presented method utilizes the approximate dynamic programming technique to iteratively solve the augmented algebraic Riccati equation on the LQT problem by using the collected information of state, reference trajectory, and input. An online algorithm is summarized.