can anybody achieve genetic algorithms by programming in mathcad

[赵亚军]

when i do some multi-objective parameter optimization cases in mathcad,i find that there are only three algorithms that can be used，L-M，Conjugate gradient method and Quasi-Newton method. these methods may can't find the global optimum,or take very long time.nowadays,there are many other algorithms,such as GA(Genetic Algorithm),it may solve these problems easily.

but my mathematics is poor，can any body do it,or has anybody done that?

i want to have a try,but my methematics is poor,at the same time,i am not proficient in programming.

wish for your help.

[ValeryOchkov]

See please - http://communities.ptc.com/docs/DOC-2565

[赵亚军]

wow,you are so intelligent

but your function don't apply to my case.i have more than 2 parameters,can you afford me your worksheet about GA program，my version is Mathcad 15.

if you can develop a model that can be used to solve all the multi-objective parameter optimization cases,that's more better.such as Minerr,minimize,a generic model，not limited to two parameters.

[ValeryOchkov · Accepted solution]

See the atach

[赵亚军]

thanks very much

it's too complex and difficult for me,but i will do my best

[ValeryOchkov]

See the picture:

[赵亚军]

you know,my english is not well,it takes too much time for me to read the book

i have another two questions:

1.how to regress the parameters with constraints? minimize and minerr have this ability,if the MGA can achieve，it may be used in future Mathcad versions.

2.the other limitation is the independent variables,here x can be a vector,but if i want x to be a matrices,how?

i would appreciate it very much if you could help me.of cause,you have helped me a lot. thanks again.

[ptc-4723985]

Good day, Valery!

Can you send me the example of programm code for MGA with some function as example?

I tryed to create the function in your file, but the Mathcad is given to me the error.

I don't know about this erro, maybe I make something wrong!

Best regards,

Kirill

[ValeryOchkov]

Kirill T wrote:

Good day, Valery!

Can you send me the example of programm code for MGA with some function as example?

I tryed to create the function in your file, but the Mathcad is given to me the error.

I don't know about this erro, maybe I make something wrong!

Best regards,

Kirill

See the picture 3.31 here http://twt.mpei.ac.ru/ochkov/Mathcad_12/Fig_Titles.html

[ptc-4723985]

Valery,

Thank you!

Kirill

[ValeryOchkov]

键 张 wrote:

thanks very much

it's too complex and difficult for me,but i will do my best

This problem has been described in this book:

http://www.springer.com/chemistry/physical+chemistry/book/978-3-7091-0530-6

[RichardJ]

these methods may can't find the global optimum

There is no algorithm that can be guaranteed to find the global optimum of a non-linear optimization problem.

nowadays,there are many other algorithms,such as GA(Genetic Algorithm),it may solve these problems easily.

Maybe. Maybe not. It depends on the problem: http://en.wikipedia.org/wiki/No_free_lunch_in_search_and_optimization

[赵亚军]

Richard Jackson 编写:

There is no algorithm that can be guaranteed to find the global optimum of a non-linear optimization problem.

as what you said,but i want to have a try,my task got stuck,the most important reason is that it takes so long time，i have tried minimize and minerr,they all failed. if this problem can't be solved,i will get in big trouble. do you have any idea to sovle the problem that the time required to find the minimum is so long?

this is only a small part of data,if i add more data,it may can't finish the caculation.what i am most concerned about is it can give a result,no matter the local optimum or the global optimum.