Research On Heave Motion Prediction Of Surface Effect Ships

Surface effect ships have been widely used in civil and military fields due to the virtues of high speed,small wave-resistance,good maneuverability,etc.When surface effect ships are engaged in offshore operations such as lightering and replenishment,the heave movement will unavoidably occur due to the influence of the waves,which seriously affects the safety of cargo replenishment between ships and increases the risk of personnel and cargo during the transfer process.If the heave motion of surface effect ships can be effectively predicted,and the appropriate control scheme is adopted in advance to offset the influence of the wave disturbance,the heave motion amplitude of surface effect ships will be greatly reduced to ensure the safety of offshore operations.Therefore,it is of great theoretical value and practical significance to study the heave motion prediction of surface effect ships.Firstly,the heave motion mathematical model of a surface effect ship and the wave disturbance mathematical model are established in this thesis.Then,under the condition that the heave position is immeasurable,and the heave acceleration can be measured by the inertial measurement unit,a method combining empirical mode decomposition with frequency domain integration is utilized to estimate the heave position according to the heave acceleration of the surface effect ship.To solve the problem of heave motion prediction caused by wave disturbance,the prediction model of the heave motion based on multilayer feedforward neural network is established.The back propagation algorithm is used to train the established neural network prediction model in accordance with the estimate of the heave position of the surface effect ship to realize the prediction of the heave motion.Further,in order to improve the prediction accuracy,the prediction model of the heave motion based on long short-term memory neural network is established.The water cycle algorithm is used to train the established neural network prediction model according to the estimate of the heave position of the surface effect ship to realize the prediction of the heave motion.Simulations on a surface effect ship are carried out for the proposed heave motion position estimate method and the designed heave motion prediction model of the surface effect ship using MATLAB at sea states one to six,respectively.The results demonstrate that the proposed heave motion position estimate method and the designed heave motion prediction model are effective.