This paper is concerned with the secure H_(∞) control problem for cyber-physical systems(CPSs)under time-delay attacks,which cause the systems’information to be received with delays.Through transforming the H_(∞) c...This paper is concerned with the secure H_(∞) control problem for cyber-physical systems(CPSs)under time-delay attacks,which cause the systems’information to be received with delays.Through transforming the H_(∞) control problem to a two-player zero-sum game problem,the H_(∞) controller can be obtained by solving the saddle point for the game.Due to the unknown system parameters,a model-free reinforcement learning(RL)method is applied,and the measurement feedback controller is designed based on the historical inputs and the tracking errors.To obviate the requirement of the initial admissible policy,the value iteration(VI)of the adaptive dynamic program(ADP)algorithm is proposed to solve the approximate saddle point.Moreover,the con-vergence of the proposed algorithm is proved.Finally,a networked circuit system is employed to show the effectiveness of the proposed methods.展开更多
基金supported by the Natural Science Foundation of China under Grant 61973277 and 62073292in part by the Zhejiang Provincial Natural Science Foundation of China under Grant LR20F030004.
文摘This paper is concerned with the secure H_(∞) control problem for cyber-physical systems(CPSs)under time-delay attacks,which cause the systems’information to be received with delays.Through transforming the H_(∞) control problem to a two-player zero-sum game problem,the H_(∞) controller can be obtained by solving the saddle point for the game.Due to the unknown system parameters,a model-free reinforcement learning(RL)method is applied,and the measurement feedback controller is designed based on the historical inputs and the tracking errors.To obviate the requirement of the initial admissible policy,the value iteration(VI)of the adaptive dynamic program(ADP)algorithm is proposed to solve the approximate saddle point.Moreover,the con-vergence of the proposed algorithm is proved.Finally,a networked circuit system is employed to show the effectiveness of the proposed methods.
