Optimal design of asymmetrically arranged moorings in a floating production system based on improved particle swarm optimization and RBF surrogate model

Abstract

In offshore oil and gas production, the increasingly complex line system and relatively fixed incidence direction of external loadings have led to asymmetrically distributed mooring lines in floating production systems (FPSs). This study presents an approach for the optimal design of asymmetrically arranged mooring systems in FPSs, considering the mooring radius, azimuth, separate angle, number of lines and three segment lengths as design variables. A series of sample mooring configurations were generated via the constrained Latin hypercube sampling method and executed by a time-domain numerical model. An improved particle swarm optimization (PSO) method involving both continuous and discrete design variables in a constrained design space was applied to mooring optimization. The radial basis function (RBF) model was trained by the sample data and applied as a surrogate model to replace the expensive finite element simulations. A case study of a deep-water semisubmersible platform was presented to demonstrate the implementation of the optimization procedure. The results showed that the improved PSO exhibited fast convergence and strong robustness. The offset of the floating platform was reduced by 8.29 % via the asymmetrical mooring pattern.