First, you should perform you experiments according to an experimental design and obtain the models (response surface model or after training based on these data you can obtain artificial neural network model). Afterwards, you should done the model-based optimization using genetic algorithm; matlab program is suitable software to perform such calculations.
see some theoretical outline on:
http://www.sciencedirect.com/science/article/pii/S0039914008004050
You should create the chromosome (solution) structure, then develop a neighborhood structure (mutation and crossover). These structures can be find in the related literature of your subject. Finally, insert these items in the process of the GA.
Dear Sayed
the most crucial issue here is how to construct the chromosome and the cost function , so you need first to define the problem clearly in order to be able to define these two things. the rest of the process can be then developed easily..
best of luck
- Badges
- Science topic