AUV rendezvous trajectories generation for underwater recovery

Abstract

-For many years, autonomous underwater vehicles (AUVs) have been developed and employed for a myriad of tasks. Their ability to accurately collect and monitor oceanic conditions makes them a valuable asset for a variety of naval missions. Deploying and recovering AUVs, however, is currently largely limited to surface vessels or swimmers. The purpose of this paper is to demonstrate that by using a mathematical technique called a direct method of calculus of variations, it is possible for an AUV to autonomously compute and execute a trajectory that will allow for recovery by a submerged mobile recovery system (another AUV, submarine, etc.). The algorithm ensures that a smooth trajectory is produced that, while not traditionally optimal, is realistic and still close to the optimal solution. Also, using this technique allows the trajectory to be computed very rapidly allowing it to be recomputed every couple of seconds to accommodate sudden changes, possible adjustments and different disturbances, and therefore to be used in the real life.

A single acoustic beacon-based positioning method for underwater mobile recovery of an AUV

“…Generally, in order to simplify a docking mission, it is assumed that the HOV is traveling in a race track at a low velocity during the rendezvous process, which consists of two long straight track legs and two semicircle track legs as shown in Figure 1. Recovery can be thought of as taking place in three stages 16 : Initial stage. When the AUV reaches around the recovery area (at position A), it broadcasts the acoustic signals by the transducer.…”

Aplicación de Muestreo basado en Modelos de Control Predictivo a un Vehículo Autónomo Subacuático

“…Kawano has a four part system [4]: 1) the offline Markov Decision Process module performs the motion planning offline; 2) the replanning module determines a path when there are new obstacles; 3) the realtime path tracking module gives action that needs to be taken; 4) the feedback control module regulates the vehicles' velocity to the target velocity. Yakimenko uses the direct method of calculus to generate trajectories that are kinematically feasible for the vehicle to traverse [5]. There are four main blocks: the first generates a candidate trajectory that satisfies boundary conditions and position, velocity and acceleration constraints; the second employs inverse dynamics to determine the states and control inputs necessary to follow the trajectory; the third computes states integrating the guidance and control via a method called shrinking horizon model predictive control [6].…”

Dubins-RRT Path Planning and Heading-Vector Control Guidance for a Uuv Recovery

“…In [3], a current observer based on kinematics constraints is designed to compensate drift caused by ocean currents. The contribution of paper [4] is to give a mathematical method of calculus of variations of a recovery path for a submerged mobile. The bio-inspired control method and the cascaded controller consisting of a kinematics controller and a dynamic controller are proposed in [5].…”