Parameter optimization for FPSO design using an improved FOA and IFOA-BP neural network

In the offshore oil industry, FPSO (floating, production, storage and offloading) units play a leading role for the production, processing and storage of oil. The hull girder strength of FPSO, which is related to the safety and economic aspects, is usually designed based on engineers’ experience. In this study, a novel method is presented to optimize the FPSO design parameters which mainly affect the hull girder strength. The proposed method employs an improved fruit fly optimization algorithm (IFOA) and IFOA-BP model which combines IFOA and back-propagation (BP) neural network. Firstly, the IFOA-BP model maps the nonlinear relations between the input and output variables, and then the reserved network can predict the stress value of critical position and the self-weight of FPSO for any set of design parameters. The numerical results indicate that the IFOA-BP model has a remarkable predication ability. Further, the reserved IFOA-BP model and the proposed IFOA is used to search for the optimal set of design parameters. Compared with the contrastive design, the optimal set of design parameters obtained using the proposed method gives lower stress value of critical position and smaller self-weight of FPSO. The optimization results show the advance and superiority of the proposed method.