In the genetic algorithm there is not strong relation between the parameters. We consider a method on the basis of RSM which first of all specifies the obtained solution by GA is near the global optimal solution or not? When the response is negative, then we choose an approach to obtain a precise solution which is closer to the optimal solution. This approach is a combination of GA and RSM, We solve some problems such as QAP and Timetabling with this approach and obtain some satisfied results.

Keywords