Optimum Design Of Mooring System Based On BP Neural Network

Ships or platforms which are working at sea must maintain a relatively fixed position. As the floating structures don't have the resilience for the horizontal movement, therefore, the mooring system is generally needed. The mooring system provides resilience so the floating structures can keep operating in a relatively fixed position under the environment load. How to reduce the movement to ensure the safety of operations, the selection and layout of mooring lines are particularly important.To solve the above issues, optimum design of mooring lines'layout which is based on the BP neural network simulation are studied. Mathematical optimization model: the azimuth of mooring lines and anchor's horizontal distance are optimization variables, the force of each mooring line can meet the strength requirements as the constraints, the horizontal movement which the probability of occurrence of wave in several directions are considered is the optimization goal. As the complexity on time domain analysis of movements, constructing explicit objective function is difficult. Furthermore, the time domain motion analysis takes a long time. As the most important characteristic of BP neural network is the nonlinear mapping function, it can approximate any nonlinear function. Therefore, BP neural network can be used to simulate the movements and force in time domain analysis. In this paper, taking a crane ship working in the South China sea as an example, the time domain analysis is done by Moses which is an authority software, it lays the foundation for subsequent neural network simulation. Different arrangement of the mooring lines are designed by orthogonal experiment which takes the azimuth of the mooring lines and anchor's horizontal distance as experimental factors. Time domain analysis for each mooring design is done by Moses. The results as the training sample are used to train the BP neural network. The optimal arrangement of mooring lines are gained by using genetic algorithm. The movements have a significant reduction after optimization. The research methods and results of this paper can be applied to similar optimization problem, it provides a reference to the layout of mooring lines for floating structures.