GENETIC ALGORITHM COUPLED WITH A DETERMINISTIC METHOD FOR OPTIMIZATION IN ELECTROMAGNETICS

In this paper, a hybrid technique for global optimization based on the genetic algorithm and a deterministic method is presented. A potentia l advantage of the hybrid method compared to the genetic algorithm is that global optimization can be performed more efficiently. An intrins ic problem of the hybrid techniques is related to the moment of stoppi ng the stochastic routine to launch the deterministic one. This is inv estigated using some natural criteria for the commutation between the two methods. The results show that it is possible to gain in efficienc y and in accuracy but the criterion is usually problem dependent. Fina lly, to show the solution of a real problem the hybrid algorithm is co upled to a 2D code based on the boundary element method to optimize a connector of 145 kV GIS.