Reinforcement learning-based receding-horizon control for optimal output regulation problem with sensor and actuator attacks.