simulation-based non-linear constrainted multiobjective optimization, any suggestions to which solver to use?

Question

Jason Xu 2019-4-12

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/455995-simulation-based-non-linear-constrainted-multiobjective-optimization-any-suggestions-to-which-solve

编辑： Matt J 2019-4-12

Hello everyone, im using FEA(finite element analysis) software to calculate some physical models, and together with matlab optimization toolbox to optimize my design. However my FEA simulation is very time-consuming, to be more specific it takes around 1 min to build the FEA model hence for each design, it takes 1min to check the constraints and 1min to evaluate the objective function. My problem is stated as followed:

1) Is it possible to build my model only once ?

2) which optimization algorithms should i use? in matlab for multi-objective only ga and pareto search are avaliable, and GA is definitely one way to go but it might takes too muhc time. I dont have a solid background in maths so thats why im here to ask for help.

Thank you.

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Matt J 2019-4-12

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/455995-simulation-based-non-linear-constrainted-multiobjective-optimization-any-suggestions-to-which-solve#answer_370351

编辑：Matt J 2019-4-12

1) See Objective and Nonlinear Constraints in the Same Function.

2) If ga takes too much time and pareto search is the only other option, there doesn't appear to be much choice. However, you should see if the UseParallel option in ga makes the runtime more tolerable for you.