optimisation agrithm

Hi, everyone

Hope you are well and thanks for your attention. Recently I frequently met a problem which might look stupid but I really want get some professional suggestion from Matlab user centre.

The problem looks like this:

There is an object function which can be denote like f(a,b,c,d,...), a,b,c,d... are variables within known range. I need to find out on which condition f reaches its maximum or minimum.

My current agrithm is to change every variable a little bit within their ranges and compare the object function. However, in my case, there are 300 variables in total which makes my life messy. So, I am wondering if there is any existed arigthm to do that or at least an easier way ?

All the best Dongsheng

Answer 1

[Fangjun Jiang]

doc fmincon

Answer 2

[Walter Roberson]

With 300 variables, you are almost certain to have many many many local maxima and minima. fmincon() as suggested by Fangjun is only a local minimizer.

There is a prerecorded webinar on Global Optimization; the files for it are available here

The introductory documentation for the Global Optimization Toolbox is over here