Novel Model-free Optimal Active Vibration Control Strategy Based on Deep Reinforcement Learning

Figure 11

Episode and average reward of DDPG during training of the partially observed MDOF system.