A Cooperative Motion Control Using Leader-Follower Approach for Unmanned Underwater Vehicle

This paper aims to address the point-to-point tracking cooperative motion control problem during the recovery process of Unmanned Underwater Vehicle (UUV) in ocean engineering. A cooperative motion control method based on the leader-follower approach and double Deep Q-Network (DQN) decision network is proposed. Line-of-Sight (LOS) guidance is combined with the design of longitudinal velocity and heading angle virtual control laws. A double DQN structure with fixed Q targets is designed to reduce the complexity of the training algorithm, thus stability and efficiency are improved. The challenge of parameter selection in UUV controller design is addressed. A controller based on the DQN decision network is proposed. It estimates the expected cumulative reward of each action based on the current state, selects the optimal action, and significantly reduces the time required for UUV to reach the desired position. Furthermore, in a disturbed environment in ocean engineering, disturbance signals are incorporated as inputs to the DQN decision network and considered during training. This enhances control accuracy and robustness in disturbed environments. Simulation results validate the effectiveness of the proposed method. A new approach is provided to improving the stability, speed of training and the accuracy of control during the UUV recovery process in ocean engineering.