OPTIMIZATION OF FLEXIBLE PIPES DYNAMIC ANALYSIS USING ARTIFICIAL NEURAL NETWORKS

摘要

Irregular wave dynamic analysis is an extremely computational expensive process on flexible pipes design. One emerging method that aims to reduce these computational costs is the hybrid methodology that combines Finite Element Analyses (FEA) and Artificial Neural Network (ANN). The proposed hybrid methodology aims to predict flexible pipe tension and curvatures in the bend stiffener region. Firstly using short FEA simulations to train the ANN, and then using only the ANN and the prescribed floater motions to get the rest of the response histories. Two approaches are developed with respect to the training data. One uses an ANN for each sea state in the wave scatter diagram and the other develops an ANN for each wave incidence direction. In order to evaluate the accuracy of the proposed approaches, a local analysis is applied, based on the predicted tension and curvatures, to calculate stresses in tension armour wires and the corresponding flexible pipe fatigue lifes. The results are compared to those from full nonlinear FEM simulation.