Obstacle avoidance based on double closed loop control of autonomous underwater vehicle for submarine cable laying

When the autonomous underwater vehicle (AUV) lays the submarine cable in complex environment, it should balance the laying quality and its own safety. In this paper it designs an engineering modeling method under the condition of irregular obstacles and an obstacle avoidance method based on double closed loop control for AUV to lay the submarine cable. The model of irregular obstacle of bulge terrain and depression terrain are established respectively. Then the outer-loop controller based on model predictive control is designed to find the optimal path of AUV in order to avoid the unexpected obstacle. The segmented progressive optimization technique is used to keep the suspension height with the seabed. It could keep the minimal change of suspension height and obtain the shortest path. The best cabling path is designed by detecting the contour feature point and distinguishing the different work situations. The inner-loop controller based on Gaussian process is designed to reduce the influence of current interference. It compensates the control input with the mean value of Gaussian process by the outer-loop controller. Also this controller could reduce the outer-loop following error efficiently. In the simulation it compares double closed loop control with sphere model predictive control and double model predictive control respectively with the different obstacles. The result shows that double closed loop control improves the quality of the submarine cable laying and AUV's safety in the submarine complex environment effectively. •When the autonomous underwater vehicle (AUV) lays the submarine cable in complex environment, it should balance the laying quality and its own safety. This paper designs an engineering modeling method under the condition of irregular obstacles and an obstacle avoidance method based on double closed loop control for AUV to lay the submarine cable. The model of irregular obstacle of bulge terrain and depression terrain are established respectively. Then the outer-loop controller based on model predictive control is designed to find the optimal path of AUV in order to avoid the unexpected obstacle. The inner-loop controller based on Gaussion process is designed to reduce the influence of current interference. And this controller could reduce the outer-loop following error efficiently.