well, fuzzy logic and neural network are artificial intelligence used in process control nowadays. So, is GA suitable in process control too?

 

Hi All,

The Genetic algorithm are more or less an optimization tool. It is used in order to define which algorithm is the better solution for a specific proble. To do this optimization, GA experiment the various efficiency of a population and after select the mos adapted and modify the less adapted in order to create a new generation. The GA execute thius cycle up to a certain
target (define as target value to maximize or minimize). In general GA are working on a mathematical model of the real problem to solve. By the fact that GA is based on experiments, it is difficult to imagine to run these
experiments on a real process. As an exemple GA is often used to tune (adapt weight factors or redesign architecture) neural network. As exemple many experiments are currently drive to optimize mobil robot conrol.

In conclusion, yes GA is suitable for process control (for control design) but not for continuous control.

Best regards

S.Monnet