Abstract: An incremental iterative Q-learning algorithm (IIQLA) is proposed to tackle the optimal secure control problem for cyber-physical systems under false data injection attacks. Within a ...