Adaptive Vibration Control of Smart Structure Using Deep Reinforcement Learning

In this research, the authors developed an adaptive control method using deep reinforcement learning which is a kind of machine to suppress vibration smart structures. This just requires information about response and input, don’t require numerical models for controlled object design controller. We experimented verify effectiveness method. experiment, structure fabricated by aluminum plate piezoelectric actuator was used as object. Three kinds algorithms are employed, Deep Q Network (DQN), Deterministic Policy Gradient (DDPG), Twin Delayed DDPG (TD3), performance compared. As result, we succeeded in reducing norm frequency impulse disturbance up 40 dB compared uncontrolled case. demonstrates applicability control.