•  
  •  
 

Abstract

The sloshing of liquid in a partially filled tank is of great concern to aerospace vehicles, road vehicles and ships due to complex nonlinear motion involving the free surface of the liquid. Sloshing is a type of violent liquid motion that is created by forces such as an impact and breaking waves. Oscillation of the liquid in the vessel may threaten navigation safety. This paper attempts to predict the hydrodynamic forces induced by the motion of liquid inside a tank based on experiment and Recurrent Neural Network (RNN). The experiment of sloshing test is carried out using the Stewart platform at Changwon National University (CWNU). The motions of the LNG carrier are inputted into the Stewart platform, then the hydrodynamic forces induced by the liquid in the tank are analyzed. Then, an RNN is used to model the complex dynamics of the sloshing load in the liquid tank and the RNN predictions are compared with the experimental results. This new approach, which offers robust modeling of the complex dynamics of a sloshing load in an LNG carrier, can be applied to simulate the ship maneuvering in waves.

COinS