Research On Dynamic Positioning Control System Based On Model Predictive Control

Most ships and offshore floating platforms are equipped with dynamic positioningsystem (DPS) in different levels.They achieve dynamic positioning without mooring systems.In order to keep target status and attitude, they use various sensors to detect their statuschanges and external environment interfere (such as wind, wave, current, etc.). Then, processall types of real-time information using their own automatic control system. Calculate thenecessary control forces which make them move to the target position and attitude. As ahigh-tech control system, in recent years, DPS achieve a good dynamic position effect in thecomplex marine environment. DPS gets more and more attentions all over the world and it iswidely used in almost all kinds of ocean structures.This paper chooses some deep-water drilling platform as the study object and designs amodel-based predictive dynamic positioning control system(Model Predictive Control，MPC).The research work is as follows:（1）A linear mathematical model of low frequency for the deep-water platform motionis established according to rigid body dynamics and kinematics theory. The environmentalinterference model of wind, current and second-order wave is also established at the sametime.（2）According to the Dynamic Matrix Control (DMC) algorithm of MPC theory,Dynamic Positioning DMC controller is designed based on the low-frequency linear motionprediction model.（3）For the characteristics of propulsion system of Deep-water drilling platform, thethrust distribution problem is analyzed especially the thrust allocation logic. Solutions ofvarious types of propulsion system constraints and limitations are proposed and foursimplified thrust allocation schemes are designed for redundant rotating propeller propulsionsystem.（4）This paper designs the semi-physical simulation system of deep-water drillingplatform dynamic positioning. Both dynamic positioning control system simulation andexperiments are implemented. The simulation is under the simulation system taking intoaccount wind, wave and current environmental effects. The results validate the feasibility and effectiveness of model-based predictive dynamic positioning control system.Deep-water drilling platform motion model is built on the basis of MPC theory. Thedynamic positioning control system is designed and the thrust allocation schemes areproposed in this paper. The simulation and experiment results show that model-basedpredictive control of dynamic positioning system is feasible and effective and it can obtaingood control and positioning effects.