fmincon optimiser becomes very slow or it gets "stuck" for quite some time at a certain point of the computation.

2 次查看（过去 30 天）

benho 2016-2-13

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/267970-fmincon-optimiser-becomes-very-slow-or-it-gets-stuck-for-quite-some-time-at-a-certain-point-of-the

编辑： benho 2016-2-17

I was given this piece of code to optimise an aerofoil. However, i discovered that the fmincon optimiser often gets stuck at certain stage of computation. The value of the first order optimality stays constant at 1.554e+04 and I have no idea how to make this better?

my tolerances are diffminchange 0.05, tolcon 0.05 and tolfun 0.05. I noticed that my first order optimality is often very high. What is actually happening to the fmincon optimiser?

2 个评论
显示无隐藏无

Matt J 2016-2-13

编辑：Matt J 2016-2-13

No way to know without seeing the code. Did you profile() the code?

Your tolerances are very high, though. I would imagine that a bigger problem will be that the solver stops too early.

benho 2016-2-17

编辑：benho 2016-2-17

drag.m

What I discovered later when i troubleshoot the code was that it kept giving me this error.

At line 85 of file ../src/userio.f (unit = 5, file = 'stdin') Fortran runtime error: End of file

i looked it up on the forums and it was something to do with how my windows computer is talking to xfoil. i've already added the lines which were provided on this page (<http://www.mathworks.com/matlabcentral/answers/79287-linking-fortran-to-matlab>) but the same error still pops up I could put the code here which is calling up xfoil. Any help would be greatly appreciated!

请先登录，再进行评论。

请先登录，再回答此问题。