Abstract
The survivability of autonomous underwater vehicles (AUV) in complex missions and dangerous situations is of great significance to ocean resource exploration, hydrological research, maritime rescue, and undersea military. Existing researches on motion control for the AUV mainly focus on its normal operating, but the active self-rescue method in emergency situations is hardly found. As classical control methods are not sufficient enough for complicated self-rescue missions of the AUV, this paper uses the deep reinforcement learning (DRL) algorithm to solve this problem because the DRL algorithm has the advantages in learning and decision making for complex robot control missions. In this paper, the normal motion control of the AUV based on the deep deterministic policy gradient algorithm is explored, including the yaw angle adjustment, yaw angle adjustment extension, trajectory tracking, and normal floating-up control of the AUV. Then, active self-rescue methods are successfully achieved to recover the AUV from emergencies, such as ocean water density decreasing sharply or one fin getting jammed at a random angle. What is more, real environment experiments are successfully conducted on a self-developed platform of the AUV to validate the feasibility of the proposed control methods. The results can effectively improve the survivability of the AUV and can be a reference to submarine survivability technologies.
Highlights
Autonomous underwater vehicle (AUV) is playing an irreplaceable part in ocean resource exploration, hydrological research, underwater military, maritime rescue, and others
This article proposes an attitude-based control strategy to reach the goals of motion control for normal operating, as well as active self-rescue for emergencies, such as ocean water density decreases sharply or one fin gets jammed at a random angle
Four agents are trained in the simulation environments: Agent 1 accomplishes the yaw angle adjustment task with a small range, Agent 2 deals with the standard floating-up task, Agent 3 carries out the floating-up task when ocean water density decreases sharply, and Agent 4 for the floating-up task when one fin gets jammed at a random angle
Summary
Autonomous underwater vehicle (AUV) is playing an irreplaceable part in ocean resource exploration, hydrological research, underwater military, maritime rescue, and others. This article proposes an attitude-based control strategy to reach the goals of motion control for normal operating, as well as active self-rescue for emergencies, such as ocean water density decreases sharply or one fin gets jammed at a random angle. The active self-rescue research of AUV is rare because its control is too difficult for variable emergency types in the hostile underwater environment. Taking the great potential of the DRL algorithm into account for the control of AUV, this paper adopts a state-of-art DRL algorithm to realize the normal motion control and active self-rescue control on a physics simulator and conducts experiments with a selfdeveloped X-rudder AUV. (3) An active self-rescue method based on the DDPG algorithm for AUV is proposed to realize the autonomous recovery in emergencies when the seawater density decreases or the fin of AUV is jammed. (1) After the agent model for the yaw angle adjustment task with a small range is trained via the DDPG algorithm, a yaw angle adjustment extension method is adopted to enable the AUV to realize the yaw angle adjustment with a large range. (2) Based on the yaw angle adjustment control strategy, a trajectory tracking control strategy is demonstrated by a sinusoids curve tracking simulation experiment. (3) An active self-rescue method based on the DDPG algorithm for AUV is proposed to realize the autonomous recovery in emergencies when the seawater density decreases or the fin of AUV is jammed. (4) A self-developed AUV is used in real experiments to verify the feasibility of the proposed control strategies
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have