The Design And Simulation Of Sail-Assisted River-Sea Direct Link Ship Heading Neural Network Control System

With the improvement of economy,the consumption of energy is growing.Because the energy used by vessels is mainly fossil fuel at present,which inevitably produce pollution problems.With the development of the state and society,more attention has been attached to the environment problems gradually,and energy conservation has been an important path for the development of society.In this context,Sail-assisted ship become one of the hot topic in the shipping industry because its economy and environmental benefits.Due to the particularity of River-sea direct link ship,its emission requirements are more stringent.he installation of sails can reduce energy consumption and reduce pollution,but the sail from the force,may have a negative impact on the ship’s heading stability.Through the establishment of the sailing of the River-sea direct link ship movement model,the design of the heading control system,simulation study to test the ship in a straight state under the conditions of different wind direction to maintain the direction.First of all,according to the characteristics of the target ship,the appropriate sail has been selected which match to the target ship.The sail size and position had been set on the premise that deck area and the ship driving and navigation has been not affected.Then the finite element method was used to analysis the aerodynamic characteristics of the sail.And the maximum thrust of the sail was obtained under different wind conditions.So the force and torque was obtained which were caused by the sail in the navigational condition.Secondly,for the ship with sail,the 20000DWT River-sea direct link ship was set as target ship,and the hull,propeller,rudder on the hydrodynamic and torque were calculated.And the role of the wind on the hull was calculated as well.Based on those data,movement mathematical model with three degrees of freedom was established which is based on the MMG model.Then the Matlab/Simulink was used to simulate the direct route and rotary simulation experiment of the target ship.Due to the complexity and non-linearity of the ship motion model,the conventional PID controller is difficult to meet the requirements.According to the nonlinearity and high steady state of the neural network,the single neuron network PID control model and the cerebellar model mixed with the PID control model were designed.And the steering gear model was set as the object controlled by those models.Then the response speed and smoothness of the output of those models compared with the conventional PID control model.The control effect of those models was evaluated according to the comparative results.The three controllers were used as the controller of the heading control system,and the parameters of each controller are set.According to the different heading deviation,the heading,steering angle,heading deviation and track of the motion model were set as output which were used to verify the feasibility of those systems.Then the ship motion model was simulated under different wind conditions.And the output of conventional PID control,single neuron network PID control and cerebellar model PID controller were compared with each other.So the merits of each controller were obtained according to the comparative results.The simulation results show that under the normal sea conditions,the three controllers can maintain the stability of the ship while the ship can draw advantage from the sail.What is more the effect of the conventional PID control and single neuron PID control is similar,and the cerebellar model PID controller is best.