Design And Application Of Data-driven Cloud Design And Optimization System

As an important branch and key technology of multidisciplinary design and optimization,data-driven design and optimization method(also known as surrogate model-based design and optimization method)has developed rapidly in recent years,and has been widely used in the field of product design and optimization.This method is based on a small number of sample points to construct an approximate model of a simulation or test mode,and the approximate model is used to replace the original model for optimization analysis.This method can effectively solve the problems of long product design cycle and high cost.Especially for the design optimization of complex systems,this method is of great significance.With the development of data-driven design and optimization theory,the corresponding design optimization toolbox and commercial software have also been greatly developed and widely used.However,these toolboxes and software still have problems such as high cost,slow technology update and complex operation.To solve these problems,this paper develops a data-driven cloud design and optimization system,referred to as DADOS.The main features of the system are: 1)Cloud-based deployment,easy to use;2)Expandable framework design,which can realize real-time technology updates;3)Process-based task building framework design,simple operation;4)Support for multi-person collaborative work.Centered on the design,development and application of DADOS system,the main research contents of this paper are as follows:Firstly,based on the data-driven design and optimization process,a design optimization toolbox of the DADOS system is built.The toolbox is composed of multiple modules such as experimental design,modeling and performance evaluation,and optimization analysis.Each module contains a variety of algorithms to meet the needs of different engineering problems.Secondly,according to the main design elements of the system,the overall framework and functional framework of the system are designed.The system adopts the MVVM mode as the overall system architecture mode,and is built using the technical framework of Vue.js + Spring Boot + Flask + My Batis,which realizes the separation design of the client and server sides of the system.The system is mainly composed of multiple functional modules such as portal and login,system management,component management,task management,etc.,which can well meet the normal optimization design requirements.The system is deployed on the cloud,and users can log in and operate directly through a browser.Thirdly,verify the stability and accuracy of the system by test function library.Through the repeated calculation of some test functions in the test function library,the DADOS system has good stability.By comparing all test functions with modeling and optimization analysis results,it is verified that the DADOS system proxy model module has good modeling capabilities,and the calculation results of the optimization analysis module are more realistic and reliable.Through the horizontal comparison of the model accuracy and optimization analysis results of all test functions,it is verified that the DADOS system surrogate model module has good modeling capabilities,and the calculation results of the optimization analysis module are realistic and reliable.Finally,based on the DADOS system,two different engineering problems are optimized and analyzed.The application of the system in two different engineering cases shows that the DADOS system has strong practicability and universal applicability.