Adaptive tracking control of underwater vehicle-manipulator systems

An adaptive control law based on a sliding mode approach for the tracking problem of underwater vehicle manipulator systems (UVMS) is developed. Based on the body-fixed and joint-space coordinates dynamic model, the proposed control law avoids the inversion of the system Jacobian, thus overcoming the occurrence of kinematic singularities. Furthermore, to avoid representation singularities of the orientation, the attitude control of the vehicle is achieved through a quaternion based error. A Lyapunov-like analysis is used to prove the error convergence. A numerical simulation presented shows the performance of the control scheme in a case study involving a full-dimensional task.