Multi-objective optimization of Tension Leg Platform using evolutionary algorithm based on surrogate model. (15th January 2018)

• Record Type: Journal Article
• Authors: Zhang, Xinshu; Song, Xingyu; Qiu, Wenzhen; Yuan, Zhiming; You, Yunxiang; Deng, Naiming
• Is Part Of: Ocean engineering. Volume 148(2018)
• Page Start: 612
• Page End: 631
• Publication Date: 2018-01-15
• Subjects: Multi-objective optimization -- TLP -- Airgap -- Tendon tension -- Radial basis function; Ocean engineering -- Periodicals; Ocean engineering; Periodicals; 620.4162
• DOI: 10.1016/j.oceaneng.2017.11.038
• Languages: English
• ISSNs: 0029-8018

Abstract

Abstract: An Innovative Tension Leg Platform (TLP) Optimization Program, called ITOP, has been developed to solve the multi-objective optimization problem for TLP. We first examine the hydrodynamic behavior of a base TLP for wave headings between 0 ∘ and 45 ∘ . The numerical results show that the maximum heave and surge motion responses occur in 0 ∘ wave heading in long-crest waves. It is found that the dynamic tension of No. 8 tendon is larger than the other tendons and reaches its maximum in 45 ∘ wave heading. It can be attributed to the fact that heave and pitch motions are almost out of phase for wave periods between 10 and 15 s. Because the maximum wave elevation occurs near the northeast column and the vertical motion is very small, the minimum airgap occurs there. Moreover, a surrogate model based on radial basis function (RBF) has been built and adopted to estimate the hydrodynamic performance of TLP. A multi-objective evolutionary algorithm, Non-dominated Sorting Genetic Algorithm II (NSGAII), is employed to find the Pareto-optimal solutions. By comprehensive and systematic computations and analyses, it is revealed that the maximum dynamic tension shows positive correlation with pontoon height and width, but negative correlation with hull draft, column spacing, and column diameter. The most efficient modification strategy for design is proposed to reduce the maximum dynamic tendon tension. According to the strategy, the column spacing, draft, and column diameter should be increased in sequence. By applying this strategy, the maximum dynamic tendon tensions can be reduced while the total weight of the platform is minimized as much as possible. Highlights: A surrogate model based on radial basis function is built and adopted to estimate the hydrodynamic performance of TLP. The optimization results are verified by direct numerical simulations to ensure the accuracy of the surrogate model. The Non-dominated Sorting Genetic Algorithm II (NSGAII) is performed for the multi-objective optimization of TLP. The most efficient modification strategy for design is proposed.