Font Size: a A A

Structure Performance Optimization And Prediction Of SPS Hatch Cover

Posted on:2020-12-17Degree:MasterType:Thesis
Country:ChinaCandidate:Z WeiFull Text:PDF
GTID:2392330590950967Subject:Ships and marine structures, design of manufacturing
Abstract/Summary:PDF Full Text Request
Steel polyurethane sandwich plate system(SPS)has been widely used in engineering because of its light weight,good energy absorption,sound insulation and noise reduction,especially in ship construction and maintenance.In this paper,the lightweight design goal is to use SPS sandwich plate instead of steel hatch cover to form the structural design scheme of SPS hatch cover.Based on the mechanical properties analysis,the response surface method based on Box-Behnken experimental design(BBD)is used to optimize the structural properties of SPS hatch cover.The structural optimization scheme of SPS hatch cover is obtained,and the Stochastic Forest prediction model is established to realize the structural forces.Accurate prediction of academic performance.The main contents of this paper are as follows:(1)The basic process of structural analysis by finite element method is introduced.By comparing the optimization design methods such as sequential quadratic programming and gradient optimization,and combining the structural characteristics of SPS hatch cover in this paper,the response surface method with high accuracy and simple operation is adopted to optimize the structure.The experimental design process,optimization process and model checking of response surface method are introduced in detail.In order to predict the mechanical properties of SPS hatch cover structure,a stochastic forest regression algorithm based on decision tree was introduced.The process of building the model and the test method of the prediction results were introduced.(2)Three design schemes of SPS hatch cover are put forward in the initial design of equal weight and the substitution of traditional steel hatch cover.Considering the processing technology,one scheme is selected as a further optimization scheme.The static strength of steel hatch cover and the optimized SPS hatch cover structure is analyzed,and the structural deformation and stress results are obtained.(3)Taking lightweight as the objective of optimization design,the response surface method based on BBD experimental design method is adopted to optimize the design of SPS hatch cover.The response surface model is established with structural deformation and equivalent stress as the objective of optimization and structural size and quality as constraints,and the optimal design scheme of weight-reducing SPS hatch cover is obtained.In order to meet the requirements of process,the optimized design scheme is adjusted.Structural size,forming two optimization schemes.(4)The static strength of the two optimization schemes is calculated,and the corresponding deformation and stress of the SPS hatch cover structure are obtained,and compared with the structure before optimization.The results show that after optimization,the structural deformation is reduced by 32.58% and 34.67%,the structural stress is reduced by 27.49% and 30.04%,the strength and stiffness meet the requirements,and the weight of the optimized scheme is reduced by 10.04% and 8.31% respectively compared with the substitution target.(5)Establish the performance prediction model of SPS hatch cover structure: using the data of SPS hatch cover designed by BBD test as sample set,using bootstrap resampling method to randomly extract multiple sub-data sets in sample set,analyze the data and establish a stochastic forest regression model on the deformation and stress of SPS hatch cover structure.(6)Accurate prediction of mechanical properties of SPS hatch cover is achieved.Random forest regression model and linear regression model are used to predict mechanical properties of SPS hatch cover structure,and the predicted results are compared and analyzed.The results show that the Stochastic Forest regression model established in this paper has high accuracy and the prediction results are more accurate.
Keywords/Search Tags:SPS hatch cover, static strength, response surface methodology, optimization, Random forest regression
PDF Full Text Request
Related items