Research On Optimization Method Of Thrust Allocation For Dynamic Positioning System Of Ships

With the increasing shortage of biological and mineral resources on land,various organizations and countries in the world have begun to look to the vast oceans and have invested a great deal of manpower and material resources in the exploration of marine resources,which to some extent has promoted the development of marine equipment technology development.Dynamic positioning system of ships is a kind of high-tech marine equipment,which is widely used in marine engineering and plays an important role in marine exploration,especially deep sea exploration.The thrust distribution link plays a crucial role in process of the dynamic positioning It is the link between the control system and the propulsion system and directly determines the size and direction of thrust to be generated by each thruster.This paper mainly researches on the thrust allcotion problem for dynamic positioning system for an engineering vessel through simulating on Matlab platform.The main tasks as follows:(1)Building the three-DOF moving mathematical model of dynamic positioning vessel,disturbance model of wind,wave,current and propulsion system.Besides,this papers designs a PID controller applied to the dynamic positioning system and verifies correctness of the vessel model and the feasibility of the designed controller by open loop simulation and closed loop simulation respectively.Moreover,building general optimization objective function and constraints equations in the case of considering the thrust error between the given thrust of the control system and the actual thrust generated by the propulsion system,the energy loss of the positioning system and the thrust limitation of the propulsion system.Based on this,a general optimization model of thrust allocation of dynamic positioning system of ships for vessel was built.(2)The adaptive genetic algorithm has the shortcomings such as premature phenomenon and the poor local search ability,and considering simulated annealing algorithm has the advantages of accepting the deterioration solution to a certain extent and better local search ability.Therefore,they are combined together to be a mixed algorithm and the adaptive genetic algorithm and simulated annealing algorithm has been presented.Then,on the one hand,the phenomenon of precociousness is better avoided,on the other hand,the ability to find the global optimal solution is also enhanced.(3)In the process of using the quadratic programming algorithm to for optimation,it often has the advantages of faster convergence rate and stronger local search ability,but it alsohas some disadvantages such as relatively easy to fall into local optimal solution and strong dependence on initial solution.Therefore,it is combined with genetic algorithm in this paper,which not only can make use of the global search ability of genetic algorithm,but also can make use of advantages of converge fast and search locally strong of the quadratic programming,it solves the problem of dependence on deficiency of original value of the sequential quadratic programming and improves the accuracy of the algorithm.(4)The adaptive genetic algorithm and two new algorithm are respectively used as optimation method of thrust allocation,which are verified by simulation and their simulation results are analyzed and compared.Meanwhile,the redundancy strategy for the thrust allocation system is simulated to verify the redundancy ability in the case of single tunnel thrust,double tunnel thrust and single azimuth thruster failure for the dynamic positioning vessel.