Control of Autonomous UnderwaterVehicles

摘要

Autonomous Underwater Vehicles find extensive applications in defense organizations for underwater mine detection and region surveillance. These are also useful for oil and gas industries in detection of leakage in the pipelines and also in many other marine industries. Underwater Robots can be categorized into two types namely (i) Remotely Operated Vehicle (ROV) and (ii) Autonomous Underwater Vehicle (AUV). A ROV is a remotely operated vehicle usually connected with the mother ship or base station through a tethered wire whereas AUV is an Autonomous Underwater Vehicle which traverses autonomously without any external interference. As opposed to ROV, control of an AUV is difficult because it is an underactuated system (whose actuator inputs are less than the number of degrees of freedom to be controlled), also the dynamics of AUV is influenced by external disturbances such as ocean current and hydrodynamic effects. The motion control problems of an AUV can be of different types such as path following, trajectory tracking, waypoint tracking and also localization. The thesis first develops path following control of a single AUV using the Serret- Frenet(S-F) frame approach and error backstepping technique. Later on the same back- stepping approach has been extended for implementation of formation control for multiple AUVs. Out of various motion control strategies, this thesis mainly focusses on path following control problem of a single AUV. To address this problem of path following, a virtual frame is considered. This virtual moving frame is called the S-F frame. The purpose of using S-F frame is to represent the AUV kinematics in terms of virtual frame parameters. Then a suitable control strategy has been developed which generates appropriate thruster force and rudder orientation enabling the AUV to follow the desired path. In the thesis, the path following controller has been developed using the concept of error backstepping method. In the developed controller it is also shown that the path following error i.e. distance between virtual frame and AUV actual frame approaches to zero and it is also ensured that other states of the AUV remain stable and bounded. Although error backstepping approach has been employed for path following problem but the earlier work [1] has not considered the surge motion dynamics and coupling of rudder angle. Therefore, this thesis has addressed the limitation of [1] and developed the backstepping controller considering the rudder coupling term. Although using a single AUV has many advantages but in case of its failure, the com- plete mission may be affected. Further, the area coverage by an individual AUV is limited. Thus, multiple AUVs are deployed for achieving a co-operative operation. Co-operative working of multiple AUVs obviate the aforesaid disadvantages as the group of AUVs in co-operative motion provides robustness in case of an individual AUV failure. Recently, a lot of research has been directed on developing cooperative motion control of multiple AUVs. Co-operative motion control can be achieved through different control strategies such as Leader-Follower, Virtual Based structure and Behavior Based Formation Con- trol. These cooperative control strategies have their own advantages and disadvantages. Hence, these strategies have been reviewed and in this work, the concept of S-F together with error backstepping approach have been exploited to develop formation control of multiple AUVs. A fuzzy logic controller has also been implemented for deriving the con- trol algorithm for leader-follower formation control scheme applied to control a group of AUVs. Subsequently, the thesis presents a graphical simulation environment using VRML and SIMULINK3D to visualize the effect of controllers developed in providing the desired path following and formation control activities of AUV(s). This graphical simulation accepts the AUV states as inputs and rep