Overall Design And Optimization Of Manned Submersible Based On Parameterization

With the development of economy and the progress of society,land resources are depleted day by day.And people pay more attention to the exploitation and utilization of marine resources.In this case,marine equipment,as an important tool for human to engage in marine development,has been developed rapidly.As an important marine equipment,manned submersible has its outstanding advantages and particularities.In recent years,its development has shown a trend of whole sea depth range,high performance,multi-function and engineering.Overall design is the key stage in the design and research of manned submersible,which has a decisive impact on its overall performance.Therefore,the research on optimization methods of overall design of manned submersible has an important impact on its overall performance and even on the progress of manned submersible technology.This paper combines the design of manned submersible with multidisciplinary optimization,and mainly carries out the following work:According to the subject characteristics of manned submersible,the subject decomposition of manned submersible is carried out,which is divided into pressure structure,resistance,propulsion and energy disciplines,and the related parameters of the overall design are determined.According to the design principle of the pressure structure,the initial design of the pressure structure is carried out on the basis of the code and the finite element method from the angles of material selection and structural form.Based on the pressure structure,the CFD method is used to complete the initial design of the resistance subject,and the theoretical comparison of the initial design results verifies the accuracy of the resistance calculation using the CFD method.The parameterized analysis process is studied and designed in the subject of pressure structure and resistance.And the sampling method of sample points in the parameterized process is analyzed.Based on Python and Java language,the macro files of structural finite element analysis software ABAQUS and computational fluid dynamics software STAR-CCM+are redeveloped respectively where realize the automatic continuous analysis and calculation of three-dimensional modeling,meshing and boundary conditions driven by parameters.The experimental design sample points needed for establishing approximate models are obtained.The subject performance analysis is carried out for the design variables of pressure structure and external resistance.The approximate model technology and commonly used approximate models are introduced.Kriging model,radial basis function neural network model and response surface model are used to fit the test design sample points of structural and resistance disciplines respectively.The test results show that the response surface model has better fitting effect.Therefore,the response surface model is used to fit the sample points of test design for the subjects of pressure structure,resistance and propulsion respectively,and the disciplinary analysis model that can meet the engineering requirements is obtained.The analytical model of energy discipline is established by using the response values and empirical formulas of propulsion discipline.The commonly used multi-objective search strategy and optimization strategy are introduced.Based on the optimization parameters of overall design and data flow of overall design optimization,the mathematical model of overall design optimization of manned submersible is established.By comparing the results of optimization with two different search strategies,the second generation non-dominated solution sequencing genetic algorithm(NSGA-II)with better approximation is adopted to obtain the final overall design optimization results.By comparing with the initial scheme,three typical optimization schemes are obtained,which provide reference for the overall design work of other manned submersibles.