Research On Thrust Allocation And Optimization Algorithm Based On Fault Tolerance Of Ship Thruster

With the further exploration and development of the ocean,the dynamic positioning system has been widely used.Ship dynamic positioning is a ship using its own thrusters to generate a certain thrust,so that the ship to maintain a certain position or course.The thrust allocation unit is an indispensable unit of the dynamic positioning system,which is responsible for the conversion of the control instructions into the specific thrust and azimuth of the thrusters.In this paper,the thrust allocation is the research direction,and the thrust allocation objective function and model as well as the thrust allocation optimization algorithm are deeply studied,and the main contents are as follows:First of all,the mathematical models of the ship are established,including the dynamic model,the kinematic model and the state space model.The marine environment model is established,including the sea breeze model,the ocean wave model and the current model.Designing the PID control system of the ship,and the simulations of the open loop and closed loop are carried out respectively,which verifies the correctness of the ship model and the PID controller.Secondly,the thrust dynamic weighting coefficient adjustment strategy is adopted,and the thrust allocation objective function with the fault tolerance of thruster fault is established.The major goal is that the ship can achieve accurate dynamic positioning after a thruster failure,meanwhile,ensuring that the remaining thrusters are not abrupt.The characteristics of the thruster are analyzed,and according to the limitation of the thruster's performance and physical conditions,the corresponding inequality constraints are given.A thrust allocation model with the fault tolerance of thruster fault is established.Thirdly,introducing the ant colony algorithm into the field of thrust allocation,and two improvements are made.The reason of the first improvement is that the basic ant colony algorithm may be fall into the local extremum and the global search speed is slow when the number of the ants is larger,and a search strategy to improve the global search speed of the algorithm is “a group of two ants together searches for a path”,and the adaptive mechanism and threshold function are introduced to improve it's global search capability,so a fast global ant colony algorithm is formed;The second improvement is based on the first improvement,because in the late stage of the algorithm,the speed of convergence become slow and the precision of solution is not enough,using the idea of simulated annealing,and the simulated annealing global ant colony algorithm is formed.Finally,the simulation experiment of thrust allocation is carried out by using three kinds of algorithms which are used to simulate when a main thruster failed or a side thruster failed of the ship.Simulation results demonstrate the correctness of the thrust allocation objective function and model with the fault tolerance of a thruster,meanwhile,it is testified that the basic ant colony algorithm and it's two improved algorithms can be used to solve the problem of thrust allocation.The simulation's results of the three algorithms are compared with each other and then analyzed,and it is proved that with the gradual improvement of the algorithm,the precision of the solution and the capability of global search are improved;Among them,the overall performance of the simulated annealing global ant colony algorithm is the best,followed by the fast global ant colony algorithm.