DQN Algorithm Based Acoustic Vibration Control of Tiny Objects on a Chladni Plate

Controlling tiny objects by acoustic vibration is a new non-invasive and non-contact control method. Its non-contact control of tiny objects has great advantages in many applications of biology, chemistry, engineering and physics. The current research is generally to observe the impact of acoustic vibration on tiny objects, but the method of controlling the movement of objects is still the traditional control method. This paper mainly uses reinforcement learning to control object movement, which can optimize the control effect. In this paper, a system for controlling the motion of tiny objects through acoustic vibration is designed to control the non-contact displacement of objects, and a tiny object movement control algorithm based on Deep Q Network(DQN) algorithm is proposed to optimize the control effect of particles. The effectiveness of the algorithm in acoustic manipulation is verified by experiments.