Multi-dimensional Taylor Network Optimal Control Of Dynamic Positioning Of Marine Oil Drilling Platform

A large quantity of petroleum resources is stored in the ocean.With the rapid increase of human demand for oil,it is of great practical significance to study offshore oil exploration technology.The traditional marine oil drilling platform adopts anchor positioning which is suitable for working in shallow water area.However,the cost of anchor chain increases and positioning accuracy decreases with the increase of water depth.Dynamic positioning method is adopted in this thesis,and its positioning accuracy is not influenced by water depth.Drilling platform may deviate from the target position when it is disturbed by the waves,wind and currents.According to the deviation between measured position and target position,control system calculates the thrust required to return the platform to the target position.Finally,thrust distribution algorithm is used to determine the thrust force and thrust direction of each thruster,so as to keep the platform in the target position.In this thesis,a control method of drilling platform based on multi-dimensional Taylor network(MTN)optimal control is proposed,and this method is compared with the traditional PID optimal control and back-stepping optimal control which needs the accurate mechanism model of the controlled object and accurate interference.Their advantages and disadvantages are given based on the simulation results.The main work of this thesis includes the following points:1.Establish a mathematical model of drilling platform.According to the need of dynamic positioning system,a three-degree-of-freedom nonlinear mathematical model which includes low-frequency and high-frequency motion model for drilling platform with surge,sway and yaw is established.Discussing the impact of wind,waves and currents on the platform and establishing mathematical models for them.2.Design Kalman filter.High-frequency movement of the platform does not change the average position and direction,and the low-frequency movement will make the platform offset.In order to reduce the loss of the thruster,eliminating the impact of high-frequency signals is needed.According to the integrated platform motion model,Kalman filter is designed to obtain the low frequency signal needed by the system.3.Design controllers.PID optimal control back-stepping optimal control and MTN optimal control not requiring the accurate mechanism model of controlled object like PID are used in this thesis.The first-order filter is introduced to solve the differential explosion problem in the derivation of virtual control item.The proportion,integral and differential of the error are the basic items to design the MTN optimal control.Three controllers' parameters are optimized by the improved simplex method.4.Thrust allocation.Thruster type and installation method are selected.Thrust distribution is regarded as a nonlinear programming problem,and its objective function and constraints are determined.Simulated annealing algorithm is used to optimize thrust distribution.5.Simulation platform design.To facilitate user interaction,a graphical user interface is developed.The anti-interference ability of the three control methods under different environmental interference is compared.The results show that in most cases,MTN optimal control has the strongest anti-interference ability,back-stepping optimal control is the second and PID optimal control is the worst.