| The dynamic positioning(DP) system is particularly important for the vessels in the deep sea operations (such as drilling platforms,floating cranes,etc.). Its main task is to make the direction of the vessel's fore and the vessel's position constant or make the vessel move along a predetermined trajectory under the influence of the propeller. The propulsion system is an important part of the vessel dynamic positioning system. The thrust allocation unit can distribute the control instructions which the control system outputs at the longitudinal, transverse and the fore's swing directions to each propeller properly and effectively, and can meet the minimum energy requirements at the same time.In recent years, the intelligent optimization algorithms are applied and developed in the thrust distribution area. The genetic algorithm is an evolutionary optimization algorithm based on reproduction, crossing and mutation. The particle swarm optimization algorithm is a group intelligence algorithm which uses the inertia weight, the learning factor and the mutation probability based on fuzzy rules to realize its function. As everyone knows,each algorithm has its own advantages and disadvantages and an optimization algorithm could not solve all the optimization problems. Therefore, this paper proposes a hybrid particle swarm optimization algorithm which has stronger robustness based on the hybrid optimization mode.The semi-submersible platform consisting of eight azimuth thrusters is taken as the research object, use the proposed hybrid particle swarm optimization algorithm and then research the thrust distribution optimization techniques of the semi-submersible platform.The simulations are all carried out at the MATLAB environment.First of all, the mathematical model of semi-submersible platform is built. The vessel's movement can be described by a movement mathematical model when under the influence of both the thrust of itself and the environmental interference forces it suffers. On the basis of selecting the desired coordinate system, the paper establishs the vessel dynamics and kinematics model and designs PID controller to simulate and verify the mathematical model.Secondly, the thrust distribution model is build and tnen modify the thrust allocation algorithm. On the basis of an outline of the thrust allocation problem, the paper gives a detailed description of the thruster layout and implementation constraints of the semi-submersible drilling platform, and then the mathematical model of the propeller is established. In order to improve the particle swarm algorithm's defects that it's easy to fall into local optima and its convergence of late stage is slow, combining the traditional cross,complex cross and mutation of the genetic algorithm into the standard particle swarm algorithm; and linearizing the inertia weight and learning factor of the algorithm; and then carring out a numerical verification of the improved hybrid particle swarm algorithm using the test data. Finally the simulation results show the effectiveness of the hybrid particle swarm optimization algorithm. After determining the target function of the semi-submersible platform, the thrust allocation process of the hybrid particle swarm optimization algorithm is designed, making the preparation for the further thrust distribution simulation.Thirdly, the thrust allocation optimization algorithm is simulated. It's simulated in two different environmental disturbing forces which represent two different sea state and the simulation results show the correctness of the hybrid particle swarm algorithm on the thrust distribution applications. The paper compare the thrust distribution of the sequential quadratic programming (SQP) algorithm and discuss the effects of the maximum number of iterations and the population size on the hybrid particle swarm optimization.Finally,a comprehensive overview of this paper is proposed,and the further research of the thrust allocation in the aspect of the dynamic positioning system is guided. |
PDF Full Text Request