Multidisciplinary Design Optimization For Deepwater Artificial Seabed System

Nowadays,dry tree and subsea tree are used for petroleum and gas development in deep water,but both have limitations:the drawbacks of dry tree system lie in the huge platform dimensions,high requirement of station keeping and so on;the drawbacks of subsea tree system are the high drilling cost,difficlut flow assurance and so on.To overcome these drawbacks,the DAS system concept has been proposed by Professor Huang Yi’team in Dalian University of Technology.Through designing the seafloor,subsurface and surface production sub-systems,DAS system overcomes the drawbacks of dry tree and subsea tree systems and has good adapatablility to harsh ocean environmental elements,such as strong wind and huge wave.DAS system is a complicated system which consists of artificial seabed,rigid riser,mooring system and flexible jumper,and its design process concerns theories and analysis methods of several disciplines,such as hydrodynamics,structure mechanics,uncertainty and economics.To obtain the entire design scheme of DAS system,it is necessary to balance and coordinate all the displines,as there is interaction among these disciplines.However,the traditional spiral design approach needs many repeated trials and revisions,which leads to the poor efficiency.In addition,it is difficult to obtain the global optimum design,due to the lack of consideration on interaction.Multidisciplinary design optimization is a design methodology for complicated systems,and it can improve the performance and shorten design cycle through adequately considering the interaction and coordination mechanism among disciplines.Consequently,this thesis aims to establish the overall optimization method of DAS system,based on the theory and method of multidisciplinary design optimization.The key isssues in the research process are the discipline analysis methods,distributed design framework of DAS system,the approximate method and optimization algorithm suitable for DAS system,the multidisciplinary design optimization of DAS system under the uncertainties of parameters and surrogate models.The main content can be concluded as:(1)Establishing the overall optimization model of DAS system.Firstly,based on the awareness of design features of the main sub-systems in DAS system,the limitations of the existing truss seastar artificial seabed are analyzed:as the support platform of subsurface production sub-system,the artificial seabed provides the top tension for mooring line and rigid riser,and it is a critical sub-system in DAS system;however,the existing truss seastar artificial seabed cannot provide the required top tenion for mooring line and rigid riser;besides that,its offset would lead to the secondary distribution of tension on the mooring line and rigid riser,which is detrimental to their mechanical performances.Due to the above problems,an innovative artificial seabed which is consisted of internal buoyancy can and outer platform is proposed to replace the truss seastar artificial seabed,thus the DAS system concept is improved.Then,the main dimensions planning process of the innovative artificial seabed is proposed,and the main dimensions,internal configuration,weight and center of gravity of the innovative artificial seabed are determined.Finally,based on the overall properties of DAS system,the relevant design parameters,criteria and overall performance parameters are identified,and the overall optimization model is formulated.(2)Discipline analysis methods and sensitivity analysis of design variables of DAS system.Firstly,the overall optimization of DAS system is divided into several disciplines,and the input and output variables of each discipline as well as the coupling relationship between disciplines are clarified.The relatively simple disciplines are merged,and three disciplines of cost,drag force and global dynamic response are presented finally.Subsequently,the drag coefficient prediction method for artificial seabed and the global dynamic response analysis method for DAS system are established based on computational fluid dynamics method and lumped mass method respectively.Moreover,the parameteric modeling and analysis technologies of the two disciplines are presented to conduct discipline analysis automatically.Finally,within the global dynamic response discipline,the sensitivity analysis of design variables is conducted by using2~k factorial design method and the presented parameteric analysis technology.The significance level of each variable is estimated through ultilizing hypothesis testing method,and the design variables are screened based on the calculation results.(3)Deterministic multidisciplinary design optimization for DAS system.Owing to the multidisciplinary design optimization process,which uses computational fluid dynamics method and lumped mass method many times for discipline analysis,would consume huge computational resources,the surrogate model technology is studied firstly.The surrogate models of discipline analysis are constructed based on second-order response surface methodology,support vector regression,radial basis function and backpropagation neural networks respectively,and the prediction accuracy,development and execution times of each surrogate model are compared.According to the results,backpropagation neural network is adopted.After that,the multidisciplinary design optimization model of DAS system is established based on the multidisciplinary feasible strategy,and the genetic algorithm,multi-island genetic algorithm,simulated annealing algorithm and particle swarm optimization algorithm are employed to solve the model respectively.The optimization efficiency,robustness and global optimization ability of each algorithm are compared,and it is found that the particle swarm optimization algorithm is the most sutiable one.Finally,considering the poor convergency and robustness of the standard collaborative optimization strategy,a group transmission based collaborative optimization strategy is proposed.(4)Uncertain multidisciplinary design optimization for DAS system.The interval model is adopted to quantify the uncertainties of design variables and surrogate models,and the solution method of interval function is established through translating the solution of interval function into two optimization problems.Considering that the inefficiency of the traditional nested double loop optimization method,an enfficient interval uncertain optimization method is proposed based on surrogate model technology,and the numerical examples are used to verify the effectiveness.On this basis,the interval uncertain optimization method and the investigation of deterministic multidisciplinary design optimization on DAS system are combined to investigate the uncertain multidisciplinary design optimization on DAS system.The results of uncertain optimization and deterministic optimization are compared,and the necessity of considering the uncertainties in the multidisciplinary design optimization of DAS system,is discussed.According to the optimization results,an overall optimization approach for DAS system based on interval model and multidisciplinary design optimization is presented.This thesis is of great significance for improving the overall performance and design efficiency of DAS system,and provides a new reference for the design of complex offshore structures such as Spar,tension leg and semi submersible platforms.