Key Structural Design Parameters Extraction And Correlation Fitting Method For Complex Products And Its Applications

In the process of complex product design,simulation and optimization analysis,the number of structural size parameters,the wide variety of parameters of the simulation process,the high dimensionality of performance parameters and the complex correlation characteristics make the computational complexity of complex products in simulation analysis and structural optimization.High,it is difficult to obtain the optimal design scheme for complex products with different performance analysis.In this paper,the electromechanical liquid-gas multi-disciplinary complex products such as tire building machine are taken as research objects,which reduces the computational complexity of complex product simulation and improves the efficiency of simulation optimization analysis.The research focuses on key structure size parameter extraction and multi-source key parameter correlation.Technical methods such as fitting and its application in structural optimization of complex products have been studied.The first chapter mainly summarizes the research status of complex product parameter sampling test method,parameter correlation fitting method and multi-objective optimization technology at home and abroad,and analyzes the problems and deficiencies of the above technical methods.The main research content,and introduced the organizational structure of this article.The second chapter proposes a method for extracting key structure size parameters of complex products based on multi-level description sampling.The multi-fork constraint tree is used to eliminate the follow-up dimension parameters.The multi-level uniform description sampling method is used to achieve uniform random sampling of the structure size parameters in the normal effective design space.Based on the sampled data,the Nataf transform is used to solve the response function statistical matrix.And use the method of correlation analysis to achieve the elimination of invalid size parameters and the further extraction of key structure size parameters,and reduce the number of structural size parameters.The third chapter proposes a multi-source key parameter correlation fitting method based on the equivalent simplified model.The parameter mapping relationship between key structure size parameters and simulation process parameters is constructed,and a multi-source key parameter set is constructed accordingly.Based on the multi-source key parameter set,the optimal Latin hypercube algorithm is used to obtain the parameter design matrix.The equivalent simplified model is constructed based on the parameter design matrix and performance parameters to realize the dimensionless fitting of multi-source key parameters.The fourth chapter proposes a structural optimization analysis method for lightweight and load-bearing performance of complex products.The response objective function is constructed based on the equivalent simplified model,and the constraints of input parameters are set.An improvement based on uniform decomposition method is proposed.The NSGA-? multi-objective optimization algorithm implements the optimization analysis of complex product structure according to this algorithm,reduces the computational complexity,and obtains a more uniform and accurate Pareto solution.The fifth chapter develops a parametric design,simulation and optimization integration platform for complex products,which realizes the extraction of key structural size parameters and the non-dimensional fitting of multi-source key parameters into equivalent simplified models,structural optimization analysis and data visualization.Used in complex products such as tire building machines.The sixth chapter summarizes the whole thesis,summarizes the main research contents and innovations of this paper,and analyzes the inadequacies of this paper and looks forward to the future research work.