Optimization-Based Multi-Attribute Decision Making for the 7th Generation Semi-Submersible Drilling Unit

The 7th generation semi-submersible drilling units (CSDU) are characteristic of deeper drilling depth, site locations and higher operational efficiency, compared with the last generation ones. Given the enormous live loads change and increasing trend of main size dimensions, considerable optimization should be deployed to achieve a balance of economy, safety and good work performance. Trial calculation and definite assessment to check whether alternative schemes meet the requirements turns out to be ineffective, for the case by case study of hydrodynamic and structure strength analysis is time consuming. In the paper, an integrated optimal design model is formulated by merging multi-objective optimization and multi-attribute decision making into one. A predesigned parametric Finite Element Analysis (FEA) structural model of CSDU is developed and validated and then coupled with detailed hydrodynamic analysis. Three mutually conflicting design objectives are arrived by hydrodynamic solutions. They are stability, hydrodynamic performance and steel consumption, which are screened to obtain Pareto optimality. The Technique for Order Preference by Similarity to an Ideal Solution (TOPSIS) is applied to incorporate these optimal attributes into decision-making process, considering all criteria in terms of quantitative stability, hydrodynamic performance and qualitative economy. The objective entropy coefficient measuring the importance of different attributes is introduced into weight selection for the purpose of avoiding non-determinacy and optional judgements. The optimal solutions are further verified with main dimensions of CSDUs in service and also could give predictive suggestion for the new CSDUs. The study provides a more objective way of benchmarking different structural schemes of CSDU by considering multiple criteria simultaneously. It is demonstrated that the proposed structural optimization model is capable of effectively and accurately determining the optimal design of CSDU.