optimization with a random constraints
Chaitanya Vichare
I am doing an optimization of random objective function using Genetic Algorithm (GA) in MATLAB.I have a random data from simulation whose RMS is required constraint.
Please help me to specify this RMS value of simulation data as a constraint in GA.
Thank You!
Rakesh Kumar
Hi Chaitanya,
If the constraint is stochastic I will suggest formulating a penalty fitness function instead of passing the constraints to GA. The GA function requires the constraints to be smooth. A penalty fitness should penalize the fitness if the RMS (of the data) is away from required constant. Can you post what you have tried already?
Rakesh
Chaitanya Vichare
Sorry I failed to watch your reply.I am still facing that problem.
I used the nonlinear constraint.I tried with constraint which is a nonlinear function of design variables but it is taking 2hrs. for GA to give answer while without constraint,it takes nearly 10min. Still I want to use the stochastic constant.
Here is in detail what I am doing.
1) I have "function" file in matlab in which GA command and its options are specified.
2) The fitness function is evaluated by the function which is nested in the same file.This file runs the model in Simulink for the design variable values passed to it by GA.Then using the result data of Simulation,this file evaluates the fitness function value and return it to GA.
3) In similar manner, I want to write one more function file nested in the same file as above two but this third file will evaluate constraint in a manner similar to fitness function file i.e. by running the simulink model and evaluating the constraint value from it and then giving it back to GA.
Also the time to run the file and getting answer required to be less, as I have to decide GA parameters by repeatably running program with different GA parameters.
If you want I can send my code
Please help me!
"Rakesh Kumar" wrote in message <icgj3g$ht8$1...@fred.mathworks.com>...