The control algorithm normally used in Dynamic Positioning (DP) Systems is based on linear control theory (proportional-derivative or linear quadratic MIMO controller), coupled to an Extended Kalman Filter (EKF) to estimate the environmental forces and wave filtering. Such controllers and estimators have problems of performance and stability related to large variations of loading (for tankers for example) or environmental conditions. The adjustment of controller gains and parameters of EKF is a complex process. Therefore, other techniques are being applied.
An investigation into the area of control of mechanical systems was made, carrying out theoretical and experimental studies involving nonlinear robust control techniques applied to dynamic positioning of floating vessels. Two robust control techniques were applied and compared: first order sliding mode control (SMC) and higher order sliding mode control (HOSM). It is known that the main drawback of SMC is the presence of high-frequency oscillations called chattering. This undesirable effect can be eliminated by using HOSM. In order to ascertain the performance of the controller under the DP system, time-domain simulations were done. Furthermore, the technique of sliding mode requires higher order derivatives of the vessel’s position signal. Therefore was developed an exact real-time differentiator, a mathematical technique used to obtain the signal derived from the position signal in real time. To validate the simulated controller, experimental tests were performed considering a small-scale model of a DP tanker.
The results confirmed the robustness of the HOSM controller, the good performance of the differentiator and the elimination of the chattering problem.
Comparison of Robust Control Techniques for Use in Continuous Stirred Tank Reactor Control
