Research On Subspace Model Predictive Control Of Dynamic Positioning Considering Tank Sloshing

With the proposal of China's "Ocean Power" strategy and the discovery of oil and gas fields in the South China Sea,China's marine engineering has gradually extended to the deep sea.Building a strong marine nation must develop marine equipment.LNG ships are marine equipment for transporting liquefied natural gas,when the tank is partially loaded,the problem of liquid sloshing in the tank will appear.The impact load caused by the sloshing can make the bulkhead structure have a greater structural response,and even affect the movement of the ship.At the same time,the environmental forces in the ocean will also interfere with the movement of the ship.To counteract these disturbances,the dynamic positioning system emerged as the key technology and core equipment in marine engineering.The dynamic positioning system has been applied to various offshore operating platforms,and its research and development are also extremely urgent and necessary.This article takes LNG ship as the research object and conducts research around its tank sloshing and dynamic positioning control system.First,an equivalent pendulum mechanical model for liquid sloshing in LNG tanks is established.The force and moment generated by the liquid sloshing in the LNG tank are deduced by using the theory of fluid mechanics,the equivalent pendulum model is established,and the parameters of the equivalent pendulum model are solved by the principle of equivalence and numerical solution.By comparing the results of finite element analysis and equivalent pendulum model,the effectiveness of the proposed equivalent pendulum model is proved.Secondly,the mathematical model of the ship's dynamic positioning system considering the tank sloshing is established.According to the requirements of the ship simulation system,the three-degree-of-freedom low-frequency mathematical models of ship were established,including ship kinematics and dynamics models.The equivalent single pendulum mechanical model of liquid sloshing is integrated into the hull dynamics model and organized into a kinematic model of the whole ship.At the same time,a mathematical model of the marine environment was established.Next,build a dynamic positioning control system based on neural network controller.Use the ship kinematics model to generate training data and train the neural network controller.Build a simulation module in Simulink to verify the performance of the neural network controller under different environmental conditions.Finally,construct a model predictive controller based on subspace identification,and introduce it into the design of ship dynamic positioning system.The subspace identification method is used to identify the parameters of the ship's state space equation.Construct the performance index function according to the state space equation and build a model predictive controller.The performance of the controller is simulated under different marine operating conditions and compared with the neural network controller.In this paper,the problem of tank sloshing is introduced into the design of the ship's dynamic positioning system,and the equivalent mechanical pendulum model of LNG tank sloshing is constructed.It is integrated into the ship's dynamic equations and considers the effect of tank sloshing on the ship's motion.A model predictive controller based on subspace identification is designed,and the effectiveness of the controller is verified under different working conditions.A simulation method of dynamic positioning system considering the effect of tank sloshing is proposed.Compared with the neural network controller,the superiority of the subspace model predictive control performance is proved.