Local Near-Optimal Control for Interconnected Systems with Time-Varying Delays

Abstract

In this paper, we develop a novel local near-optimal control (LNOC) for continuous-time (CT) time-varying delayed interconnected systems by observer-critic structure based on adaptive dynamic programming. A boundary function which is approximated by local state observer is presented to substitute the unknown interconnection. Through constructing a novel local value function, the LNOC is obtained by solving the local Hamilton-Jacobi-Bellman equation with the implementation of local critic neural network. By using the Lyapunov’s direct method, the proposed LNOC ensures the closed-loop CT time-varying delayed interconnected systems to be asymptotically stable. The effectiveness of the proposed LNOC scheme is illustrated using numerical simulation.