Study On Control And Thrust Allocation Of Vessels With Dynamic Positioning System

Along with the lack of land resources and the increasing needs to resource consumption,the ocean has become the focus of attention of many countries.And with the development of science technology,the exploration,development and utilization of marine resources have also gradually extended from shallow sea to deep sea.Dynamic positioning and trajectory tracking have become a hot research focus in ship engineering because of the needs to oceanic resources exploitation.As an important part of dynamic positioning system,thrust allocation takes a significant impact in improving the precision of dynamic positioning and reducing the consumption of energy.In this thesis,thrust allocation of dynamic positioning system is studied,namely how to configure thrust amplitude and direction to minimize power consumption and mechanical wear.Firstly,a dynamic mathematical model of 3 DOF,the earth coordinate system,the carrier coordinate system and the transformation principle between each coordinate system have been introduced.A oceanic disturbance in ocean is established.The basic performance of thruster is studied,and the thrust parameters can be chosen according to the propulsion performance reference diagram.In order to solve the problem that the thrust cannot be measured accurately,the relationship between thrust and revolution is obtained based on the model of thrusters.So the actual thrust can be controlled by adjusting the revolution of propulsion.Secondly,the model of thrust allocation is introduced by analyzing several constraints including singularity,mechanical wears,deviations,etc.The nonlinear objective function with constraints is obtained.Lastly,two traditional thrust allocation algorithms are discussed in this thesis.By considering their drawbacks and limitations,an improved genetic algorithm is applied to obtain thrust allocation optimum.In order to maintain population diversity and improve convergence speed,a modified fitness function and a modified mutation operator are proposed in this thesis.Simulation result shows that this algorithm improves the drawbacks mentioned above in a certain extent.