宋学官

个人信息Personal Information

教授

博士生导师

硕士生导师

性别:男

毕业院校:东亚大学

学位:博士

所在单位:机械工程学院

学科:机械设计及理论

办公地点:大方楼8021#

电子邮箱:sxg@dlut.edu.cn

扫描关注

码过的文

当前位置: 宋学官 >> 科研 >> 码过的文

A MATLAB GUI Toolbox for Surrogate-Based Design and Optimization

点击次数:

论文类型:会议论文

发表时间:2019-01-01

收录刊物:CPCI-S

页面范围:103-106

关键字:Design and Optimization Toolbox; Graphical User Interface; Design of Experiments; Surrogate Model

摘要:The surrogate model technique, a fast-solving and data-based technology, has great significances in design and optimization of complex engineering systems and has been used to approximate problems without explicit functions. In this paper, a surrogate-based design and optimization (SBDO) Graphical User Interface (GUI) named SBDO Toolbox is proposed, which combines the powerful mathematical analysis capability of MATLAB with the superb interfacial design function of GUIDE. This toolbox consists of three modules, i.e., design of experiments (DOE), construction of surrogate models and prediction/optimization. Users can address an optimization problem easily by inputting the information of independent variables, setting some options (or just keeping default values), and the results of each module can be displayed graphically. The data transmission between modules is continuous, so the integrity of design process can be guaranteed. The proposed SBDO Toolbox can be commonly applied to engineering design and optimization by simply setting parameters without any professional knowledge about surrogate model techniques. Therefore, the development of SBDO Toolbox is to simplify the steps and operations of the SBDO process in engineering systems and to promote the SBDO technology.