MODEL REFERENCE ROBUST CONTROL FOR MARINE PROPULSION SYSTEMS WITH MODEL UNCERTAINTY CAUSED BY HULL DEFORMATION

摘要

A marine propulsion system can be strongly coupled with the hull deformation in a large-scale ship. The interaction of this coupling dynamics increases the nonlinearity and complexity of the control model. Traditional control method based on the empirical PID (proportion-integration-differentiation) parameters has difficulty in dealing with the model uncertainty caused by the hull deflection. However, up to date, limited work has been done to address the model uncertainty problem of the marine propulsion system. It is therefore imperative to develop feasible and effective control systems that take the model uncertainty into account. In this study, a new method based on the model reference adaptive system (MRAS) has been proposed for the active and accurate control of the marine propulsion system coupling with the hull deformation. A finite element model of the ship shaft line was established to investigate the uncertainty boundaries of the propulsion system. Moreover, the sea trial was carried out on the hydraulic dredge named "Changjing 2" to measure the main engine power loss under hull deformation. The finite element analysis (FEA) and shipboard measurement results showed a fiducial interval of [0.1% 10%] for the model uncertainty of the marine propulsion system. These uncertainty boundaries were added into the ship speed control system model, and the MRAS controller was designed based on the Lyapunov theory to mitigate the adverse effects of model uncertainty. The stability of the MRAS has been proven by the Lyapunov stability criterion. Numerical evaluations using MATLAB®/ Simulink® software for the "Changjing 2" ship engine parameters have showed high effectiveness of the MRAS structure for speed tracking under the hull deflection coupling. This study has demonstrated that the newly proposed control system can work stably with various ship operating conditions, and its performance is superior to the traditional PID controller.