How to solve the optimization using Matlab
Información
La pregunta está cerrada. Vuélvala a abrir para editarla o responderla.
Mostrar comentarios más antiguos
How can i solve the following optimization using Matlab optimization tool box?..should i choose fmincon for this?
maximize X = a Y + b Z, where 0<a<=1, 0<b<=1 0<x<=1 and y>0
4 comentarios
Torsten
el 22 de En. de 2016
Why do you restrict the outcome of your objective function (x) to 1 if you want to maximize x ?? (I guess that x and X as well as y and Y are the same ?)
What is Z ?
Best wishes
Torsten.
Sobin cc
el 22 de En. de 2016
Torsten
el 22 de En. de 2016
Since Z is unbounded, the maximum X-value is Infinity (the problem is unbounded).
Best wishes
Torsten.
Sobin cc
el 22 de En. de 2016
Respuestas (1)
John D'Errico
el 22 de En. de 2016
Editada: John D'Errico
el 22 de En. de 2016
You don't understand. IT IS UNBOUNDED. Set Y to any value you wish. as long as b is positive, then let Z approach infinity. X will be made arbitrarily large.
If, as in a comment I see, you decide to give a limit to x as a maximum... (I can only assume that you meant X, not x since there is no x in the problem.) Still, use of fmincon is a waste of time. Just use grade school algebra and some common sense.
Pick ANY values for a, b, and Y. ANY VALUES. Then we have
X = 500 = a*Y + b*Z
So just solve for Z, as a function of a,b,Y.
Z = (500 - a*Y)/b
Since a and Y are bounded and both quite small, both bounded between 0 and 1, Z will NEVER be negative in that result. NEVER.
There is absolutely no reason to bother using fmincon here. Never forget to use common sense, replacing it with a computational algorithm. If you do, you will too often get garbage for results, and you won't know when to trust those results.
La pregunta está cerrada.
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!