Adaptive control of underwater robotic manipulators

Abstract

It is difficult to anticipate the range of dynamic effects that may be imposed on a robotic manipulator during sub-sea operations. The inherent buoyancy, added mass and viscous drag terms due to operating in water will produce different dynamic response than obtained from the arm in air. Also external disturbance loads may be imposed due to current, tides or vortex shedding at higher arm traverse rates. The magnitude of these effects could well detune an arm set up with fixed controller gains. In this paper, a self-tuning PID controller is implemented for underwater robotic manipulators based on a pole-placement design method. A second order system dynamic model with its general discrete-time transfer function is assigned to each joint of the robotic manipulator. The controller has the same form as the PID, whose digital counterpart is acquired using Tustine's bilinear transfer. An approximate recursive least squares identification method, which reduces computation time is used to identify the process parameters online. The PID controller's gains are continually updated using the most recent estimation of the dynamic behaviour of the robotic manipulator. Symbolic calculation is carried out beforehand to reduce the computational burden. A new explicit relationship between the process parameters and the controller gains is obtained for the first time. It is concluded that real-time implementation of the adaptive controller presented is possible.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>