optimization- profile for the optimized values

Dear all,

I have been working with optimization using ga for my system. I have 4 input ariables required by the system which has to be obtained from the optimizer results. Right now my optimized results look like this:

The values are within the bounds, but i want a profile with less fluctuations and also i want a increasing profile. ie,., the value of var1 at (t) > var1 at (t-1).

a. How to achieve this? Should i create a constraint matrix with my specifications?

b. Also is it possible to obtain a profile like the following as my optimal output? (0-20 represents time in mins).

Please give any suggestions for the same.

Thanks

Answer 1

[Naveen Venkata Krishnan]

Hello Nivedhitha,

Ga assumes very random intial values for the variables in the search space and moves towards optimility with the help of crossover and mutation. By profile do you mean the profile of optimization variables ? . If yes, I dont think you can get an increasing profile by using ga. But can try implementing steepest descent optimization method to get such profile.

[Nivedhitha S]

Yes by profile, I meant the profile of optimization variables. Thank you so much for your suggestion. I 'll try implementing steepest descent algorithm and get back to you.