A multi-task fuzzy control for underwater welding manipulator

Manipulators mounted on unmanned submarines are inherently kinematically redundant. Such systems, which are normally controlled by operators from within a vessel on the sea, can follow a desired path both by moving its manipulator or the submarine itself. Due to this redundancy one might control the system such that it tracks the desired path and fulfil some secondary tasks at the same time. Due to different inertia of vehicle and manipulator it is convenient to perform fast manoeuvres by manipulator and leave the submarine uncontrolled until the manipulator reaches its physical limits or some other secondary tasks are violated. This paper presents a method based on fuzzy logic for redundancy resolution of such a manipulator system during a welding mission under the sea. The method accounts for velocity disturbances induced in submarine by underwater stream. It also makes it possible to prioritize secondary tasks and also the motion in different degrees of freedom of the system.