Abstract: Artificial muscles are advantageous owing to dynamic stiffness; however, the drawback of this is their poor controllability. Here, a reinforcement learning-based control system is proposed ...